Kinematics of the equine axial skeleton -...

Kinematics of the equine axial skeleton during aqua-treadmill exercise

Submitted for the Degree of

Doctor of Philosophy At the University of Northampton

2017

Jessica I. York

© Jessica I. York, 2017

This thesis is copyright material and no quotation from it may be published without

proper acknowledgement.

Acknowledgements

This thesis is the culmination of several years hard work which has been made possible by the support of several key people. Thank you to the continued support and advice of my supervisor Dr Wanda McCormick and to interim supervisor Dr James Littlemore for stepping in at a crucial moment and offering such helpful and positive feedback. Thank you to Professor Ian Livingstone, my Director of Studies, for always being so matter of fact and reassuring, and making everything seem so simple! Thank you to the staff and students of Moulton College for their support and assistance with data collection. Particular thanks to the staff of the Equestrian Centre and Equine Therapy Centre, past and present. Thank you to the horse owners for allowing their horses to take part. Thank you to the Thomas Harrison Trust at Moulton College for funding the project and thank you to the Graduate School of the University of Northampton for their supportive and engaging research department with their endless workshops and training opportunities. Thank you to Dr Catherine Fritz at the University of Northampton for pointing me in the right direction with the statistical analyses. Thank you to Dr John Abbot MRCVS of Town & Country Veterinary Practice for his support with the horses undertaking the trials and continual interest in the project. Thank you to the Royal Veterinary College for the very generous loan of equipment used for data collection and to Dr.-Ing. Thilo Pfau for his help and advice. Thank you to Emily Sparkes of the Royal Veterinary College for her help and advice with software, data and equipment. Thank you to my PhD family of friends that have been through this tempestuous journey with me and supported me along the way; Emily, Lauren, and Clare. Thank you to my family for their unfailing support and understanding, always. Especially to my husband who has always believed in me. And finally, utmost and heartfelt thanks to Dr Anna Walker. Without whose ideas, contacts, advice, knowledge and experience this project would not have been possible. Thank you for your constant motivation, support and endless positivity. I would not have reached this point without you. Anna, thank you.

Abstract

Equine aqua-treadmills are increasingly applied within the industry for rehabilitation from injury and training. Research into aqua-treadmill exercise has been increasing yet there are still opportunities to further quantify the effect of water depth on locomotion in order to optimise aqua-treadmill protocols to see improved rehabilitation from injury or to exercise horses more effectively for their chosen discipline. Much of the current aqua-treadmill literature focusses on the effects of water on locomotory parameters in walk, thus providing an opportunity for investigation into the impact of water at trot. Trot is the favoured gait for effective quantification of lameness and symmetry studies so it was anticipated that there may be opportunities to make comparisons to previously published overground data. Effective schooling of horses overground includes training aids, such as side reins, to constructively develop a horse’s way of going and often to assist the horse in maintaining concentration. This provided a further opportunity for investigation into the use of side reins during aqua-treadmill exercise. This project therefore, aimed to quantify the effect of increasing water depths on pelvic and withers movement of horses trotting on an aqua-treadmill and to analyse the impact the use of side reins has on these movements. Seventeen sound horses were habituated to aqua-treadmill exercise and subjected to one of two exercise protocols where data were collected either by optical motion capture (Qualisys©) or an inertial sensor system (Xsens©). The exercise protocol involved trotting on the aqua-treadmill at four increasing water depths, that of the third phalanx (P3), mid fetlock, mid third metacarpal (MC3), and mid carpus. Markers for optical motion capture were located on the poll, withers (T4/T5), mid thorax (T13), tuber sacrale, left and right tuber coxae, left and right tuber ischia. Inertial sensors were located on the poll, withers (T4/T5), mid thorax (T13), lumbar vertebrae (L4), tuber sacrale, left and right tuber coxae, and top of the tail (1st coccygeal vertebrae). Data were cut into strides with accelerations double integrated to generate displacement amplitudes, both vertical and mediolateral, for statistical testing. Pitch and roll data from the inertial sensors was also extracted and processed for analysis. Data were processed using custom written scripts (Matlab®) and repeated measures ANOVAs were performed throughout to test for significance with post hoc analysis where appropriate. Water depth was found to have a significant effect on vertical displacement amplitudes of the pelvis and withers with vertical displacements increasing with increasing water depth, and a greater displacement in the pelvis than the withers. Minimum and maximum positions of the pelvis and withers were found to decrease and increase accordingly with increasing water depth, with minimum values decreasing significantly indicating an increase in limb compression during stance. Maximum vertical positions also increased significantly indicating greater maximum lift out of the water as a result of the increased compression. Water depth had no effect on symmetry of horses trotting on an aqua-treadmill and no effect on pitch amplitudes. Vertical displacements, pitch and symmetry were not altered with the addition of side reins, suggesting that the adoption of a different head and neck position whilst reaching comparable displacement amplitudes encourages

further engagement of back muscles possibly providing stimulus for building greater strength through muscular development. Water depth was found to have no effect on mediolateral displacements of the pelvis or withers but with the withers exhibiting larger mediolateral displacements than the pelvis at lower water depths but reducing to an amount comparable to the pelvis at deeper depths suggesting that deeper water provides a stabilising effect on the front end of the horse. Side reins had no effect on mediolateral displacement amplitudes or on roll amplitudes. Mediolateral flexions of the spine were not affected by water depth or side reins, suggesting that the horse can be worked harder at greater water depths without over stressing the mediolateral capabilities of the spine. Vertical displacements of the pelvis were significantly increased when trotting on the aqua-treadmill in a very low depth of water compared to measurements overground but this effect was not seen in the withers suggesting the front end of the horse can efficiently compensate for water depth by flexing at the carpus, although larger pitch amplitudes were reported at the withers suggesting a change in head and neck position to create a ‘jump up’ over the water. Side reins were found to decrease vertical displacement amplitudes in the withers overground but trotting on the aqua-treadmill in a small amount of water counteracted this effect suggesting that the addition of water may counteract a ‘downhill’ effect seen in horses wearing side reins overground. This project suggests that the aqua-treadmill is beneficial at increasing the workload for the horse that may possibly have a corresponding effect of increasing muscle mass, strength and condition, but without detrimental effects to cranial-caudal or mediolateral symmetry patterns and that side reins have a potential benefit in supporting these locomotory patterns. Knowledge of this primary scientific data will better assist professionals working with aqua-treadmills to more effectively benefit the horses with which they work. There is, however, an opportunity for further longitudinal research to further support the effective application of the aqua-treadmill as a tool for rehabilitation and training.

Publications and Conferences

Publications: York, J., McCormick, W. D. and Walker, A., (2016) Vertical Displacement of the Equine Pelvis and Withers During Trot on an Aqua Treadmill. Equine Veterinary Journal, 48 (S49), 34-35. York, J. and Walker, A., (2014) Vertical Displacement of the Equine Pelvis When Trotting on an Aqua Treadmill. Equine Veterinary Journal, 46 (S46), 55. Conferences: York, J. (2016) The kinematics of the equine axial skeleton when exercising on an aqua-treadmill. Symposium presented to: 5th Postgraduate Research Symposium, Moulton College, Northampton, 15 December 2016. York, J., McCormick, W. D. and Walker, A. (2016) Vertical displacement of the equine withers and pelvis during trot on an aqua-treadmill. Seminar Presentation presented to: International Conference on Canine and Equine Locomotion (ICEL 8), London, 17-19 August 2016. York, J., McCormick, W. D. and Walker, A. (2016) The kinematics of the equine axial skeleton when exercising on an aqua-treadmill. Seminar Presentation presented to: University of Northampton Postgraduate Conference, University of Northampton, Northampton, 14 June 2016. York, J. (2015) The effect of the aqua-treadmill on the kinematics of the equine axial skeleton. Symposium presented to: 4th Postgraduate Research Symposium, Moulton College, Northampton, 11 December 2015. York, J., and Walker, A., (2014) Vertical displacement of the equine pelvis when trotting on an aqua-treadmill. Seminar Presentation presented to: International Conference on Equine Exercise Physiology (ICEEP 9), Chester, UK, 15-20 June 2014. York, J. (2014) The use of the aqua-treadmill in equine rehabilitation and exercise. Symposium presented to: 3rd Postgraduate Research Symposium, Moulton College, Northampton, 16 December 2014. York, J. (2013) The equine aqua-treadmill and its use in equine therapy and rehabilitation. Symposium presented to: 2nd Postgraduate Research Symposium, Moulton College, Northampton, 13 December 2013. York, J. (2012) The aqua-treadmill and its use in the rehabilitation of equine lameness. Symposium presented to: 1st Postgraduate Research Symposium, Moulton College, Northampton, 18 December 2012. Posters: York, J., McCormick, W. D. and Walker, A. (2015) Vertical displacement of the equine withers and pelvis when trotting on an aqua-treadmill. Poster Presentation presented to: University of Northampton Postgraduate Poster Competition, University of Northampton, Northampton, 13 May 2015. York, J., (2014) Walking in water: Equine hydrotherapy. Poster Presentation presented to: University of Northampton Postgraduate Poster Competition, University of Northampton, Northampton, 07 May 2014.

i

Table of Contents

Table of Contents ............................................................................................... i Table of Tables .................................................................................................. v Table of Figures ............................................................................................... vii Abbreviations List ............................................................................................. x

CHAPTER 1: Introduction .................................................................................... 1

1.0 Introduction .................................................................................................. 1 1.1 Equine gait ................................................................................................... 2

1.1.1 Gait parameters ...................................................................................... 3

1.1.2 Gaits........................................................................................................ 3

1.1.3 Symmetry and lameness ......................................................................... 6 1.2 Gait analysis ................................................................................................ 8

1.2.1 Optical Motion Capture ........................................................................... 9 1.2.2 Videography .......................................................................................... 10

1.2.3 Force plates .......................................................................................... 11 1.2.4 Inertial Measurement Units ................................................................... 12

1.2.5 Treadmill locomotion ............................................................................. 15 1.3 Hydrotherapy ............................................................................................. 17

1.3.1 Human Hydrotherapy ............................................................................ 17

1.3.2 Animal Hydrotherapy............................................................................. 18 1.3.3 Equine Hydrotherapy ............................................................................ 19

1.3.4 Modes of equine hydrotherapy .............................................................. 20 1.3.4.1 Swimming horses ............................................................................ 21

1.3.4.2 Cryotherapy ..................................................................................... 23 1.3.4.3 Aqua-treadmills and walkers ........................................................... 25

1.4 Current research into aqua-treadmills ..................................................... 27

1.4.1 Heart rates ............................................................................................ 28 1.4.2 Physiological responses ........................................................................ 30

1.4.3 Muscle ................................................................................................... 32 1.4.4 Gait and locomotory parameters ........................................................... 34 1.4.5 Other studies ......................................................................................... 37

1.5 Conclusion ................................................................................................. 38 1.6 Opportunity for Research ......................................................................... 39

CHAPTER 2: Methods ........................................................................................ 41

2.0 Methods ...................................................................................................... 41

2.0.1 Subjects ................................................................................................ 41 2.0.2 Protocol ................................................................................................. 42

2.1 Use of Qualisys© ....................................................................................... 45 2.1.1 Subjects ................................................................................................ 45 2.1.2 Qualisys© Protocol ............................................................................... 45

2.2 Use of Xsens© ........................................................................................... 48 2.2.1 Subjects ................................................................................................ 48

ii

2.2.2 Protocol ................................................................................................. 48 2.3 Data Processing ........................................................................................ 54

2.3.1 QTM (Qualisys©) .................................................................................. 54 2.3.2 MTX (Xsens©) ...................................................................................... 54

2.4 Reliability .................................................................................................... 55

2.5 Validating Xsens© against Qualisys© ..................................................... 62 2.5 Statistical analyses ................................................................................... 69 2.6 Health and Safety and Ethics ................................................................... 69

CHAPTER 3: Vertical displacements of the equine pelvis and withers when trotting on the aqua-treadmill at increasing water depths .............................. 71

3.0 Introduction ................................................................................................ 71 3.1 Aims ............................................................................................................ 76

3.2 Methods and Statistical Analysis ............................................................. 77

3.2.1 Statistical Analysis ................................................................................ 78

3.3 Results........................................................................................................ 79 3.3.1 Vertical Displacement of the Pelvis and Withers ................................... 79 3.3.2 Percentage Change in Displacement of the Pelvis and Withers ........... 83 3.3.3 Symmetry .............................................................................................. 88

3.3.4 Changes in Position .............................................................................. 91 3.3.5 Percentage Change in Position ............................................................. 95

3.3.6 Pitch ...................................................................................................... 99 3.4 Discussion ............................................................................................... 101 3.5 Conclusion ............................................................................................... 108

CHAPTER 4: The Effect of Side Reins on the Vertical Displacements of the Pelvis and Withers when Trotting on an Aqua-Treadmill at Increasing Water Depths ................................................................................................................ 110

4.0 Introduction .............................................................................................. 110

4.1 Aims .......................................................................................................... 114 4.2 Methods and Statistical Analysis ........................................................... 115

4.2.1 Statistical Analysis .............................................................................. 116 4.3 Results...................................................................................................... 116

4.3.1 The effect of side reins on vertical displacements of the pelvis and withers ......................................................................................................... 118 4.3.2 Percentage change in vertical displacements of the pelvis and withers with and without side reins ........................................................................... 120 4.3.3 The effect of side reins on symmetry of vertical displacement amplitudes ..................................................................................................................... 122

4.3.4 Impact of side reins on pitch ............................................................... 125 4.4 Discussion ............................................................................................... 127

4.5 Conclusion ............................................................................................... 133 CHAPTER 5: Mediolateral displacements of the equine pelvis and withers when trotting on the aqua-treadmill ................................................................ 135

5.0 Introduction .............................................................................................. 135

iii


5.2.1 Mediolateral displacement amplitudes ................................................ 139 5.2.2 Roll amplitudes ................................................................................... 140 5.2.3 Mediolateral flexions ........................................................................... 141

5.2.4 Statistical Analyses ............................................................................. 143 5.3 Results...................................................................................................... 143

5.3.1 Mediolateral Displacement Amplitudes ............................................... 144 5.3.2 Roll Amplitudes ................................................................................... 149 5.3.3 Mediolateral Flexions .......................................................................... 152

5.4 Discussion ............................................................................................... 154 5.4.1 Mediolateral displacement amplitudes ................................................ 154 5.4.2 Roll amplitudes ................................................................................... 156 5.4.3 Mediolateral flexions ........................................................................... 158

5.5 Conclusion ............................................................................................... 159

CHAPTER 6: A comparison of overground and aqua-treadmill locomotion ............................................................................................................................ 160

6.0 Introduction .............................................................................................. 160


6.2.1 Statistical Analysis .............................................................................. 163 6.3 Results...................................................................................................... 164

6.3.1 Overground Vertical Displacements .................................................... 164

6.3.1.a Overground Vertical Displacements – Symmetry ............................. 168 6.3.2 Comparison between overground and ATMP3 vertical displacements 170 6.3.3 Pitch Angles Trotting Overground ....................................................... 173

6.3.4 Comparison between overground and ATMP3 Pitch Angles ........... 176 6.3.5 Mediolateral Displacements at the pelvis, withers and poll and when trotting overground ....................................................................................... 179

6.3.6 Comparison between overground and ATMP3 mediolateral displacements (pelvis, withers and poll) ....................................................... 181

6.3.7 Roll of the pelvis, withers and poll when trotting overground .............. 185 6.3.8 Comparison between overground and ATMP3 roll amplitudes ........ 188

6.4 Discussion ............................................................................................... 192

6.4.1 Overground vertical displacements ..................................................... 192 6.4.2 Overground mediolateral displacements ............................................. 193 6.4.3 Overground roll ................................................................................... 194 6.4.4 Effect of water on vertical displacements ............................................ 196 6.4.5 Effect of side reins .............................................................................. 197

6.4.6 Pitch .................................................................................................... 198 6.4.7 Effect of water on mediolateral displacements .................................... 199

6.5 Conclusion ............................................................................................... 200 CHAPTER 7 Discussion and Conclusions ...................................................... 201

7.0 Introduction .............................................................................................. 201 7.1 Effects of water depth ............................................................................. 202

iv

7.1.1 Overground comparisons .................................................................... 209 7.2 Effects of side reins ................................................................................ 212

7.3 Conclusions ............................................................................................. 214 7.4 Future research and limitations ............................................................. 216 7.5 Implications and applicability to the equine industry .......................... 220

REFERENCES .................................................................................................... 222

v

Table of Tables

Table 1.1: Common parameters used in gait analysis and their definitions. ................................. 4 Table 2.1: Aqua-treadmill Qualisys© exercise protocol ............................................................... 47 Table 2.2: Aqua-Treadmill Xsens© exercise protocol .................................................................. 53 Table 2.3 Vertical Displacement Reliability Data ......................................................................... 56 Table 2.4 Pitch Reliability Data .................................................................................................... 57 Table 2.5 Mediolateral Reliability Data ....................................................................................... 58 Table 2.6 Roll Reliability Data ...................................................................................................... 59 Table 2.7 Absolute Position Reliability Data ................................................................................ 60 Table 2.8 Mediolateral Flexions Reliability Data ......................................................................... 61 Table 2.9 Reliability and validation of Xsens© against Qualisys© using one test horse ............ 65 Table 3.1: Post hoc analysis of the significant main effect of pelvis versus withers on vertical

displacements of the pelvis and withers when trotting on the aqua-treadmill at increasing water depths.. ............................................................................................ 81

Table 3.2: Post hoc analysis of the significant main effect of water depth on vertical displacements of the pelvis and withers when trotting on the aqua-treadmill at increasing water depths. ............................................................................................. 81

Table 3.3: Post hoc analysis of the significant simple main effect of pelvis versus withers on percentage change in vertical displacements of the pelvis and withers when trotting on the aqua-treadmill at increasing water depths.. .................................................... 85

Table 3.4: Post hoc analysis of the significant simple main effect of water depth on percentage change in vertical displacements of the pelvis and withers when trotting on the aqua-treadmill at increasing water depths........................................................................... 87

Table 3.5: Post hoc analysis of the significant main effect of pelvis versus withers on symmetry of vertical displacements of the pelvis and withers when trotting on the aqua-treadmill at increasing water depths........................................................................... 89

Table 3.6: Post hoc analysis of the statistically significant main effect of pelvis versus withers when analysing the mean minimum part of the stride of 8 horses trotting on an aqua-treadmill at increasing water depths........................................................................... 92

Table 3.7: Post hoc analysis of the statistically significant main effect of water depth when analysing the mean minimum part of the stride of 8 horses trotting on an aqua-treadmill at increasing water depths........................................................................... 92

Table 3.8: Post hoc analysis of the statistically significant main effect of water depth when analysing the mean maximum part of the stride of 8 horses trotting on an aqua-treadmill at increasing water depths........................................................................... 93

Table 3.9: Post hoc analysis of the significant main effect of water depth when analysing the mean percentage change in the minimum part of the stride of 8 horses trotting on an aqua-treadmill. ............................................................................................................ 96

Table 3.10: Post hoc analysis of the significant simple main effect of water depth when analysing the mean percentage change in the maximum part of the stride of 8 horses trotting on an aqua-treadmill at increasing water depths. ...................................................... 97

Table 4.1: A comparison of the mean vertical displacement values (in mm) for the pelvis and withers obtained from the amalgamation of both Qualisys© (Q) and Xsens© (X) data collection trials (Q+X, n=17) and then just the values obtained from the Xsens© data collection trial (X, n=10)............................................................................................. 117

Table 5.1: Post hoc analysis of the significant simple main effect of pelvis versus withers on mediolateral displacement amplitudes of 10 horses trotting on the aqua-treadmill at increasing water depths. ........................................................................................... 146

vi

Table 5.2: Post hoc analysis of the significant simple main effect of water depth on mediolateral displacement amplitudes of the pelvis and withers of 10 horses trotting on the aqua-treadmill at increasing water depths......................................................................... 147

Table 5.3: Post hoc analysis of the significant simple main effect of pelvis versus withers on roll amplitudes of the pelvis and withers when trotting on an aqua-treadmill at increasing water depths.............................................................................................................. 150

Table 5.4: Post hoc analysis of the significant simple main effect of side reins on roll amplitudes of the pelvis and withers when trotting on an aqua-treadmill at increasing water depths. ....................................................................................................................... 150

Table 6.1: Post hoc analysis of the significant simple main effect of pelvis versus withers on vertical displacements in horses trotting overground (n=10) ................................... 166

Table 6.2: Post hoc analysis of the significant simple main effect of side reins on vertical displacements of the pelvis and withers in horses trotting overground (n=10). ...... 166

Table 6.3: Post hoc analysis of the significant simple two-way interactions and simple simple main effects for the comparison in vertical displacements of the pelvis and withers of horses trotting overground and on an aqua-treadmill at water depth of mid P3.. .. 171

Table 6.4: Post hoc analysis of the significant simple main effect of left versus right on the pitch of the pelvis or withers in horses trotting overground (n=10). ................................. 174

Table 6.5: Post hoc analysis of the significant simple main effect of pelvis versus withers on the pitch of the pelvis or withers in horses trotting overground (n=10) ......................... 174

Table 6.6: Post hoc analysis of the significant simple two-way interactions and simple simple main effects for the comparison in pitch amplitudes of the pelvis and withers of horses trotting overground and on an aqua-treadmill at water depth of mid P3.. .. 177

Table 6.7: Post hoc analysis of the significant simple main effect of OG versus ATMP3 on mediolateral displacements of the pelvis, withers and poll of horses trotting overground and on an aqua-treadmill at water depth of mid P3. ............................ 182

Table 6.8: Post hoc analysis of the significant simple main effect of anatomical location on mediolateral displacements of the pelvis, withers and poll of horses trotting overground and on an aqua-treadmill at water depth of mid P3. ............................ 183

Table 6.9: Post hoc analysis of the significant main effect of anatomical location on the roll of the pelvis withers and poll of horses trotting overground (n=10) ............................ 186

Table 6.10: Post hoc analysis of the significant simple main effect of overground versus ATMP3 on roll amplitudes of the pelvis, withers and poll of horses trotting overground and on an aqua-treadmill at water depth of mid P3. ....................................................... 189

Table 6.11: Post hoc analysis of the significant simple main effect of anatomical location on roll amplitudes of the pelvis, withers and poll of horses trotting overground and on an aqua-treadmill at water depth of mid P3. ................................................................. 190

vii

Table of Figures

Figure 2.1: An illustrative view of the forelimb of the horse identifying the four positions on the limb that were measured to define the four water depths. ....................................... 44

Figure 2.2: A horse on the aqua-treadmill illustrating the location of the light reflective hemi-spherical markers. ....................................................................................................... 47

Figure 2.3: Visual representation of the placement of the inertial measurement units of Xsens©. ..................................................................................................................................... 50

Figure 2.4: A horse fitted with the Xsens© inertial sensors prior to the start of the trial. ........... 50 Figure 2.5: A horse on the aqua-treadmill with the Xsens© sensors attached undertaking the set

exercise protocol. ........................................................................................................ 52 Figure 2.6: Mean (±SEM) vertical displacement amplitude (mm) of the pelvis (left) and withers

(right) for the test horse in both the Qualisys© and Xsens© trials at the water depth of mid P3 for both left diagonal pair (green) and right diagonal pair (pink). .............. 68

Figure 3.1: Graphical representation of the vertical displacements achieved by the os sacrum and the left and right tuber coxae (LTC or RTC) during a single stride from a sound symmetrical horse cut from left hind stance. ............................................................. 73

Figure 3.2: Mean (±SEM) vertical displacement (in millimetres) of the equine pelvis (left) and withers (right) at increasing water depths when trotting on an aqua-treadmill at increasing water depths for both left diagonal pair (green) and right diagonal pair (pink) (n=17) ................................................................................................................ 82

Figure 3.3: Mean (±SEM) percentage increase in vertical displacements of the equine pelvis (left) and withers (right) when trotting on an aqua-treadmill at increasing water depths for both left diagonal pair (green) and right diagonal pair (pink) (n=17). ........................ 86

Figure 3.4: Mean (±SEM) difference between the mean vertical displacements for left and right diagonal for the equine pelvis (purple) and withers (yellow) when trotting on an aqua-treadmill at increasing water depths with raw measurement of mm (left) and percentage change in displacement (right) (n = 17). .................................................. 90

Figure 3.5: Mean (±SEM) position (mm) in the absolute displacement of the equine pelvis (left) and withers (right) from the minimum starting point (at left hind stance) and maximum vertical point of the stride cycle when trotting on an aqua-treadmill at increasing water depths (n = 8). Left minimum vertical position (min left) (blue). Right minimum vertical position (min right) (light blue). Maximum left vertical position (max left) (green). Maximum right vertical position (max right) (light green). ........... 94

Figure 3.6: Mean (±SEM) percentage change in position (%) from the baseline water level of mid P3 of the equine pelvis (left) and withers (right) from the minimum vertical starting point (at left hind stance) and maximum vertical point of the stride cycle when trotting on an aqua-treadmill at increasing water depths (n = 8). Left minimum vertical position (min left) (blue). Right minimum vertical position (min right) (light blue). Maximum left vertical position (max left) (green). Maximum right vertical position (max right) (light green). ................................................................................ 98

Figure 3.7: Mean (±SEM) pitch of the pelvis (left) and withers (right) throughout a stride cycle when trotting on an aqua-treadmill at increasing water depths for both left diagonal pair (green) and right diagonal pair (pink) (n = 10). .................................................. 100

Figure 4.1: Mean (±SEM) vertical displacement (in millimetres) of the equine pelvis (left) and withers (right) at when trotting on an aqua-treadmill at increasing water depths both without and with side reins for both left (green) and right (pink) diagonal pair (n=17). No side reins indicated by dots. Side reins indicated by dashes. ............................. 119

Figure 4.2: Mean (±SEM) percentage change in vertical displacements of the equine pelvis (left) and withers (right) when trotting on an aqua-treadmill at increasing water depths

viii

without and with side reins for both left (green) and right (pink) diagonal pair (n=10). No side reins indicated by dots. Side reins indicated by dashes.. ............................ 121

Figure 4.3: Mean (±SEM) difference between the mean vertical displacements following left and right hind stance for the equine pelvis (purple) and withers (yellow) when trotting on an aqua-treadmill at increasing water depths with raw measurement of mm (left) or considering the percentage change in displacement (right) both without AND with side reins (n=10). No side reins indicated by dots, side reins indicated by dashes...124

Figure 4.4: Mean (±SEM) pitch of the pelvis (left) and withers (right) throughout a stride cycle when trotting on an aqua-treadmill at increasing water depths both without and with side reins for both left (green) and right (pink) diagonal pair (n = 10). No side reins indicated by dots. Side reins indicated by dashes. ................................................... 126

Figure 5.1: Orientation of the axes in a triaxial inertial measurement unit. ............................... 139 Figure 5.2: An illustration of how the anatomical landmarks of the withers (T4/5), mid back (T13)

and pelvis (tuber sacrale) form a straight line and at left hind stance (LHS) there is flexion to the left which changes to a flexion to the right as the horse changes to a right hind stance (RHS). ............................................................................................. 142

Figure 5.3: Mean (±SEM) mediolateral displacement amplitudes (in millimetres) of the equine pelvis (left) and withers (right) when trotting on an aqua-treadmill at increasing water depths both without side reins (orange) and with side reins (green) (n=10).. 148

Figure 5.4: Mean (±SEM) roll amplitudes (in degrees) throughout a stride cycle of the inertial measurement units stationed at the pelvis (left) and withers (right) when trotting on an aqua-treadmill at increasing water depths both without side reins (orange) and with side reins (green) (n=10).. ................................................................................. 151

Figure 5.5: Mean (±SEM) change in mediolateral angle of the equine spine (Angle XY, Withers - T13 - tuber sacrale) when trotting on the aqua-treadmill at increasing water depths (n=6)........................................................................................................................... 153

Figure 6.1: Mean (±SEM) vertical displacement amplitudes (in mm) of the equine pelvis (left) and withers (right) when trotting in a straight line overground as informed by the Xsens© data (n=10), both with and without side reins on left (green) and right diagonal (pink) stride pair. No side reins indicated by dots. Side reins indicated by dashes. .......... 167

Figure 6.2: Mean (±SEM) difference (mm) in vertical displacement amplitudes between left and right diagonal pair in the pelvis and withers without side reins (orange) and with side reins (green) (n=10). .................................................................................................. 169

Figure 6.3: Mean (±SEM) vertical displacement amplitudes (in mm) of the equine pelvis (left) and withers (right) when trotting in a straight line overground and when trotting at a water depth of mid P3 on the aqua-treadmill as informed by the Xsens© data both with and without side reins on the left (green) and right (pink) diagonal stride pair (n=10). No side reins indicated by dots. Side reins indicated by dashes. ................ 172

Figure 6.4: Mean (±SEM) pitch (in degrees) of the equine pelvis (left) and withers (right) when trotting in a straight line overground as informed by the Xsens© data both with and without side reins on the left (green) and right (pink) diagonal stride pair(n=10). No side reins indicated by dots. Side reins indicated by dashes. ................................... 175

Figure 6.5: Mean (±SEM) pitch amplitudes throughout a stride cycle (in degrees) of the equine pelvis (left) and withers (right) when trotting in a straight line overground and when trotting at a water depth of mid P3 on the aqua-treadmill as informed by the Xsens© data, both with and without side reins on the left (green) and right (pink) diagonal stride pair (n=10). No side reins indicated by dots, side reins indicated by dashes..178

Figure 6.6: Mean (±SEM) mediolateral displacement amplitudes (in mm) of the equine pelvis, withers and poll when trotting in a straight line overground as informed by the Xsens© data both without side reins (orange) and with side reins (green) (n=10)... 180

Figure 6.7: Mean (±SEM) mediolateral displacement amplitudes of the equine pelvis, withers and poll (in mm) in trot both overground and on the aqua-treadmill at a water depth

ix

of mid P3 both without side reins (orange) and with side reins (green) as informed by the Xsens© data (n=10)............................................................................................. 184

Figure 6.8: Mean (±SEM) mediolateral roll amplitudes (in degrees) of the equine pelvis, withers and poll when trotting in a straight line overground both without side reins (orange) and with side reins (green) as informed by the Xsens© data (n=10).. ..................... 187

Figure 6.9: Mean (±SEM) roll amplitudes of the equine pelvis, withers and poll (in degrees) in trot both overground and on the aqua-treadmill at a water depth of mid P3 both without side reins (orange) and with side reins (green) as informed by the Xsens© data (n=10). ............................................................................................................... 191

x

Abbreviations List

ANOVA analysis of variance ATM aqua-treadmill ATMP3 aqua-treadmill at a water depth of mid P3 bpm beats per minute C1 cervical vertebrae 1 etc. c. circa DIP distal interphalangeal joint (coffin joint) EMG electromyography Hz Hertz IMU inertial measurement unit km/h kilometres per hour L1 lumbar vertebrae 1 etc. L or R left or right diagonal stride pair (measured from left or right hind stance) MC3 third metacarpal (cannon bone) mm millimetres m/s metres per second MTP metatarsophalangeal (fetlock) OG overground P3 third phalanx PIP proximal interphalangeal (pastern) P pelvis S1 sacral vertebrae 1 etc. SD standard deviation SEM standard error of the mean SHJ scapulohumeral joint (shoulder) T1 thoracic vertebrae 1 etc. Ts Tuber sacrale (os sacrum) (croup) T4 / T5 thoracic vertebrae 4 and 5 (withers) W withers

1

CHAPTER 1: Introduction

1.0 Introduction

Equine aqua-treadmill exercise has become increasingly popular as a mode of

rehabilitation and training for horses due to the ability to standardise and monitor

many more variables than traditional overground training, thereby potentially being

able to deduce the exercise load. Aqua-treadmills for horses in a crude form were

first described in the literature by Auer in 1989, where the idea of the combined

use of a treadmill and a whirlpool with jets of water positioned at strategic locations

was proposed to be the ideal system for rehabilitation of a previously injured horse

(Auer, 1989). Current day demands on all sport horses require them to run faster,

jump higher and have more expression, all of which have implications on how

horses are bred and trained for specific optimal characteristics for different

disciplines. Horses are also required to remain competitive for longer putting

further strain on body systems. However, equine aqua-treadmills as a tool for both

rehabilitating and training horses with much positive anecdotal evidence of the

influences on biomechanics and subsequent performance, are still relatively

understudied in the literature to provide quantitative evidence to support these

positive anecdotal claims. The equine aqua-treadmill may be a useful tool in

helping to fulfil current day demands.

2

1.1 Equine gait

The gait of the modern equid is not reported to have undergone significant

changes since the first documentation of Equus caballus (Linnaeus, 1758) and the

first time that equine gait was measured and recorded by Eadweard Muybridge in

the 1870s. Equines, as terrestrial prey mammals, have a terrestrial quadrupedal

gait that has evolved to enable them to flee from predators at speed. The major

evolutionary changes from Eohippus through to Equus in relation to the movement

of the horse resulted in an increased length of leg, particularly the lower limb with

no muscle mass below the carpus or tarsus (Goody and Goody, 2000). This

resulted in long, lightweight limbs with a heavy hoof that acts as a pendulum to

increase the swing of the leg increasing the length of the stride.

The equine is not unusual as a quadrupedal mammal in terms of its biomechanics

in that it can alter its gait according to the terrain. There are four main gaits

associated with the horse with increasing speed and decreasing duty factor (see

Table 1.1) from walk through trot, canter and gallop, where each gait can be

described quite specifically according to the fundamentally different mechanisms,

foot fall sequence, forces, powers and metabolic costs (Minetti, 1998). There are

also a couple of rarer and more unusual gaits in the horse of the tolt and pace that

are seen in the gaited horse breeds such as the Icelandic Pony and Tennessee

Walking Horse. The gait of the equine can be broken down into specific

parameters such as displacements and the duty factor in order to describe the

linear and temporal gait characteristics. Recent application of accelerometers has

been used to identify footfall sequence and timings to quantify equine gaits further

3

to establish whether gaits should be considered as discrete entities or a continuum

(Robilliard et al., 2007).

1.1.1 Gait parameters

In order to describe, quantify and investigate equine gait, clear parameters have

been defined that necessitate consistent use to enable comparative, objective

studies between subjects, environments and observers. Common parameters are

defined in Table 1.1.

1.1.2 Gaits

Walk is the slowest, and possibly the most complex of the equine gaits with 4

beats that have large overlap times between the stance phases of the limbs and

no period of suspension (Back and Clayton, 2007). Walk speeds in dressage

horses have been reported as 1.37 m/s in collected walk to 1.82 m/s in extended

walk with the speed change attributed to a lengthened stride (Clayton, 1995). In

walk, each individual limb has a longer stance phase than swing phase i.e. a duty

factor greater than 0.5 (Biewener, 1983; Hoyt et al., 2006) which is possible due to

the periods of triple support which describes the part of the stride where three

limbs are in simultaneous contact with the ground.

4

Table 1.1: Common parameters used in gait analysis and their definitions.

PARAMETER DEFINITION

Stance Phase The portion of the limb motion cycle when the limb is in contact with the ground (Back and Clayton, 2007).

Swing Phase The portion of the limb motion cycle when the limb is free from contact with the ground (Back and Clayton, 2007).

Stride A complete cycle of the repetitive series of limb movements that characterize a particular gait (Back and Clayton, 2007).

Stride Duration The time required to complete one stride (Back and Clayton, 2007).

Duty Factor

The duration of the stance phase of a specified limb as a proportion of the total limb cycle duration or stride duration (Back and Clayton, 2007). Stride duration, stance duration and protraction duration together determine the duty factor (the fraction of the stride for which the limb maintains contact with the ground surface) from which the peak vertical force can be estimated (Witte et al., 2004).

Ground Reaction Forces (GRF) (Kinetic data)

The force of the ground against the limb that acts in opposition to the force exerted by the limb against the ground (Back and Clayton, 2007). A method for evaluating the accuracy of predicting peak ground reaction force from the duty factor using hoof mounted accelerometers to detect foot on and foot off has been identified as being accurate in the trot and symmetrical gaits while a correction factor is required to compensate for the difference between the lead and non-lead limbs of a pair in asymmetrical gaits (Witte et al., 2004). Vertical GRF is of the greatest magnitude most directly measuring specific limb weightbearing and sensitivity in grading lameness (Weishaupt, 2008). Craniocaudal GRF quantitates forces affecting forward progression—braking (deceleration) and propulsion (acceleration). Mediolateral GRF has the smallest amplitude, so few studies have used this variable (Hodgson et al., 2014).

5

Trot is a symmetrical contralateral two-beat gait which is the equivalent to running

or hopping in bipeds. In trot, the horse moves the limbs in diagonal pairs. Footfall

timings have been quantified as equal between left and right hind limbs so the trot

can be considered as symmetrical (Robilliard et al., 2007). The trot has a shorter

stance phase, a longer swing phase and the addition of an aerial phase. Trotting is

generally faster than walk with ridden trot speeds ranging from 3.2m/s in collected

trot, 3.6m/s in working trot, 4.5m/s in medium trot and 4.9m/s in extended trot

(Clayton, 1994a) while trotting in hand has been reported to be approximately

3.9m/s (Galisteo et al., 1998) plus with the addition of the aerial phase and

reduction in stance phase, results in a duty factor of less than 0.5 (Biewener,

1983; Hoyt et al., 2006).

Canter is the term used for a slow, collected gallop (Hildebrand, 1977). Canter and

gallop refer to the same asymmetrical gait at different speeds with canter being a

slower three-beat gait and gallop being a four-beat gait performed at the fastest

speeds (Back and Clayton, 2007). The footfall pattern of a horse at left lead

canter, is as follows: first the right hindlimb, then the left hind and right forelimbs as

a diagonal pair, and finally the left forelimb followed by a short suspension phase

with all limbs of the ground (Back et al., 1997). Canter has been described at

speeds ranging from 3-11 m/s from a collected canter of a mounted horse

overground to an extended canter of an unmounted horse on a treadmill (Deuel

and Park, 1990; Clayton, 1994b; Corley and Goodship, 1994), whereas gallop has

been measured over a range of speeds from 9 to 17 m/s in field conditions (Witte

et al., 2006).

6

1.1.3 Symmetry and lameness

The symmetrical gaits of walk and trot (where movement patterns are the same on

both sides of the horse) are difficult to quantify as there is always a degree of

asymmetry in any animal whether it is a biped or a quadruped (Back and Clayton

2007). The asymmetric gaits of canter and gallop are not used when assessing

gait symmetry. Asymmetry is due in part to laterality or natural ‘sidedness’

(Fredricson et al., 1980; Deuel and Lawrence, 1987; Drevemo et al., 1987) in the

same way that humans are either left or right handed. An additional influence on

laterality in horses is the standard procedure of handling horses from one side in

preference of the other. Laterality causes difficulties in defining a threshold to be

used to aid discrimination between natural asymmetry and asymmetry associated

with lameness.

Gait is often measured in horses in order to assess lameness for which trot is the

most commonly used gait (Buchner et al., 1994a). Trot is the most symmetrical

gait with emphasized flexion/extension and vertical displacements, where a sound

diagonal pair can be used as a control for the lame diagonal pair. Vertical

displacements are commonly studied when assessing lameness in horses.

However, care must be taken not to misinterpret asymmetry found during trot

when determining if a forelimb or a hind limb is lame as forelimb lameness has

been shown to cause a ‘false’ lameness to appear in the hind limbs and vice versa

due to compensatory load redistribution (Kelmer et al., 2005). Trot can also be

controlled in terms of speed with handlers being able to maintain pace with the

horse when conducting overground in hand studies. This ease of control enables a

good number of repeatable strides to be recorded, likewise trotting on a treadmill

7

or ridden trot maintains consistency of the stride when compared to in hand

trotting. Quantitative analysis of lameness commonly involves the calculation of

symmetry ratios, which refers to the calculation of a ratio aimed at objectively

quantifying describing parameter asymmetry between the left and right phases of

a symmetrical trot stride (Walker et al., 2010). Kinematic studies have calculated

symmetry ratios to enable asymmetry to be quantified numerically for analysis

between horses and conditions (Peham et al., 1996; Pourcelot et al., 1997;

Keegan et al., 2001, Keegan et al., 2004; Audigie et al., 2002; Keegan, 2007;

Church et al., 2009; Thomsen et al., 2010a; Keegan et al., 2011).

To address the low levels of inter-observer agreement, quantitative analysis of

lameness is required when making judgements on symmetry and lameness

(Keegan et al., 1998; Arkell et al., 2006; Fuller et al., 2006; Hewetson et al., 2006;

Keegan et al., 2010; Thomsen et al., 2010b; Dyson, 2011; Rhodin et al., 2013;

Hammarberg et al., 2016; Rhodin et al., 2017). This low rate of agreement

requires that objective quantitative assessment of gait can be made not only in

laboratory studies but in the field in particular during Veterinary assessments,

which may aid in pre-clinical detection of analysis, and early diagnosis and

therefore treatment may reduce time out of work and recovery time and costs and

increasing longevity of high level performance.

8

1.2 Gait analysis

Technological advances have enhanced the ways in which gait can be

investigated both in lab based studies and in the field. When measuring gait, it is

important to understand the differences between kinetics and kinematics. Kinetics

is the study of internal and external forces, energy, power and efficiency involved

in the movement of a body (Back and Clayton, 2007). Kinetic analysis is

concerned with the forces that initiate and alter motion requiring identification of

the external forces acting on a system together with their points of application and

lines of action (Jones, 1988). Kinematics is the branch of biomechanics that is

concerned with the description of movement (Back and Clayton, 2007).

Kinematics deals with linear and angular displacements, velocities and

accelerations without regard to the forces producing the motion (Jones, 1988).

Many methods of gait analysis utilise some kind of marker affixed to the skin. Skin

mounted markers can introduce errors in data due to skin movement artefact in

addition to the movement of underlying soft tissue (Clayton and Schamhardt,

2007). Often it is possible to use joint centres for affixing a marker but this relies

on the competences of the person affixing the marker to accurately locate and

palpate the joint centre, repeatably, and in a variety of subjects (Leach and Dyson,

1988). Markers are best placed on joints or bony landmarks with limited skin

movement (Van Weeren et al., 1992a, 1992b). Skin is known to move over bony

landmarks, up to as much as 12cm in proximal parts of the limb particularly at the

elbow (humeroradial joint) and stifle (tibiofemoral joint) (Clayton and Schamhardt,

2007). However, skin markers still provide a non-invasive and useful assessment

of equine locomotion.

9

1.2.1 Optical Motion Capture

Optical motion capture (mocap) (e.g. Qualisys (QTM) ProReflex, Qualisys Ltd,

Gothenburg, Sweden) is long recognised as the gold standard in gait analysis and

has been used to demonstrate symmetry of back movement in healthy horses

(Audigie et al., 1999), to study three-dimensional kinematics of the equine spine

during walk, trot and canter (Faber et al., 2000; Faber et al., 2001a, b), to show

that chiropractic manipulation results in more symmetrical pelvic motion (Gómez

Álvarez et al., 2008) and that small asymmetries might be related to subtle

lameness (Gómez Álvarez et al., 2007). It has also been used for other equine

research such as quantifying hoof deformation (Burn and Brockington, 2001) and

studying mandibular movements of chewing (Bonin et al., 2007).

Optical Motion Capture works using sophisticated infra-red cameras that reflect

light from light-reflective hemispherical markers affixed to usually bony anatomical

landmarks of an animal to reconstruct the shape of the animal in three-

dimensions. However, the sophisticated technology is very much restricted to use

in laboratories due to the large and bulky nature of the equipment and the

requirement of having to maintain consistent light qualities. Also, the cost of the

equipment restricts it to usually only enough cameras to be stationed around a

treadmill so that data from many consistent strides can be collected rather than on

an outside track where there may only be enough cameras to collect data of ten

strides or so. This produces another potential problem of the changes that are

documented to occur to gait when exercising on a treadmill rather than overground

(Buchner et al., 1994a). However, optical motion capture has been demonstrated

10

to reliably capture the small ranges of motion in the equine back that are otherwise

difficult to assess (Faber et al., 2000; Faber et al., 2001a, 2001b, 2001c).

1.2.2 Videography

Videographic gait analysis would seem the most accessible method, especially

when skin fixated markers are used, as there are many widely available pieces of

software that enable comprehensive, semi-automated tracking and analysis in

both two and three dimensions. However, when dealing with videographic

systems there are many more complications to consider that vary between the

systems such as: placement of the video camera ensuring it is precisely

perpendicular to the plane of interest which restricts how many strides may be

visible in the field of view, calibration of the frame, digitization, transformation,

smoothing and normalization of the data (Back and Clayton, 2007, chp.3).

High speed video cameras are generally less accessible but are very useful for

studies of short duration events or rapid movements as they can record at rates of

up to 10 million Hz (Xing et al., 2017), whereas a standard camcorder may record

at only 25-30 Hz. However, one study compared sampling rates of 60 and 1000

Hz and determined that with no differences in the measured parameters of the

equine stride, stance and swing duration greater than 3.3 metres per second a

camera that records at a rate of 60Hz is adequate for kinematic gait analysis in the

walk, but that faster gaits require a greater sampling rate for increased accuracy

(Linford, 1994). Quantifying movement symmetry with video filming in front or

behind the horse also requires complex post processing as the subject is

11

constantly moving towards or away from the camera thus making the calibration

and calculation of measurements more difficult and prone to errors.

1.2.3 Force plates

Kinetic analysis of gait refers to the measurements of the forces involved, for

which force plates, force-measuring horse shoes and strain gauges can be used.

Force plates are restricted to lab-based studies and are limited to providing data

for only one or two strides at a time with the added difficulty in making sure the

horse steps consistently in the middle of the plate to prevent artefacts. Also, it is

difficult to gauge the effects of different surfaces. This has resulted in the

development of a horseshoe for measuring three-dimensional ground reaction

forces in horses moving overground and across different surfaces (Frederick and

Henderson, 1970; Barrey, 1990; Ratzlaff et al., 1990; Roepstorff and Drevemo,

1993; Kai et al., 2000; Roland et al., 2005; Chateau et al., 2009). Early force

measuring shoe models were unable to provide a complete description of three-

dimensional ground reaction forces and moment vectors, and the 2005 version by

Roland et al. (2005) used strain gauge technology but was reported to be too

heavy at 860g to accurately assess gait without itself causing changes to the gait.

The most recent development of a dynamometric device using piezoelectric

sensors (Chateau et al., 2009) has been reported to accurately measure three-

dimensional ground reaction forces at walk and trot and on different surfaces, and

more recently in ridden horses, canter, circling, jumping and when measuring grip

on different surfaces (Robin et al., 2010; Camus et al., 2012; Crevier-Denoix et al.,

2014).

12

1.2.4 Inertial Measurement Units

Inertial measurement units have advanced enormously over recent years. Simple

hoof mounted accelerometers with a small size and mass have been shown to not

adversely affect locomotion (Witte et al., 2004; Pfau et al., 2006) and accelerations

of the withers when integrated through velocity have been shown to provide

displacement scores comparable to optical motion capture (Pfau et al., 2005).

However, the drawback of simple accelerometery devices is the absence of their

own recording capabilities requiring the use of at least a simple data logger using

MP3 recorders (Parsons and Wilson, 2006) or passing wires up the limbs to a

recording device attached to the trunk or distal limb (Witte et al., 2004, Pfau et al.,

2006).

Inertial measurement systems have been developed which incorporate tri-axial

accelerometers, gyroscopes and magnetometers. Xsens© is one example of such

a system (Xsens©, MTx; Xsens, Enschede, The Netherlands). This sophisticated

small and lightweight system has been developed to provide accurate

measurement of movement in all directions and rotations. Due to the high

accuracy and repeatability of the data recorded with little impact of environment,

they have been recruited for thorough investigation of locomotion outside of

laboratory conditions. Studies have identified that these inertial measurement

units allow data to be collected during unrestricted movement giving up to six

degree of freedom information (both orientation and displacement) (Keegan et al.,

2004; Pfau et al., 2005; Pfau et al., 2006; Pfau et al., 2007). Along with the double

integration of accelerations to provide displacement data (Pfau et al., 2005),

symmetry related parameters such as Energy Ratios can also be calculated for

13

each stride (Keegan et al., 2004; Pfau et al., 2007). These inertial measurement

systems have been utilized to document and quantify vertical head, trunk and

pelvic movements along with movement symmetry of sound horses being lunged

on hard and soft surfaces where information has previously been sparse (Walker

et al., 2010; Starke et al., 2012a; Pfau et al., 2012; Pfau et al., 2016b; Greve et al.,

2017).

One study specifically compared data from the inertial measurement units to that

of the gold standard of motion capture and determined that accuracy of

dorsoventral and mediolateral displacement data calculated from inertial

measurement units mounted along the spine of horses trotting over ground lay

within ± 4–8 mm (± 2 SD) of the gold standard of motion capture, and Audigié

energy ratios varied by ± 0.03 (± 2 SD) between the two methods (Audigié et al.,

2002; Warner et al., 2010). More recently, a smartphone (Apple iphone6) was

tested against the IMU based dedicated gait analysis system of Xsens© for

assessment of pelvic (a)symmetry and limits of agreement were similar enough to

suggest that smartphones could be utilized by first opinion practitioners as a

convenient alternative to the specialist more expensive devices (Pfau and Weller,

2017). Accuracy as a percentage of the range of movement was found to be in

the range of 9–21% (dorsoventral) and 10–24% (mediolateral). Considering that

experienced clinicians have been evidenced not to be able to consistently detect

asymmetries below a level of 25% (Parkes et al., 2009) therefore suggested that

the ease of use of the inertial units along with the accuracy and objectivity of the

data coupled with the potential to collect a large amount of simultaneous data

renders the inertial unit system a valid technique to collect objective limb and back

movement data during overground locomotion (Warner et al., 2010).

14

Since the initial development and validation of inertial measurement units for gait

analysis, other inertial measurement systems have been developed ranging from

simple accelerometery devices to the sophisticated Xsens© system described

above. Inertial measurement units have been used more and more in the literature

for investigating a variety of locomotor related areas, such as demonstrating that

inertial sensor-based systems can objectively assess the response to proximal

hindlimb flexion tests (Marshall et al., 2012). Inertial measurement units (Xsens©)

have been used to accurately quantify hoof on/hoof off timings with the inertial

measurement units mounted on the distal metacarpus or distal metatarsus and

sacrum (Olsen et al., 2012). Another 2012 study reported that a single pelvis-

mounted inertial sensor was accurate in quantifying hind limb foot contact timings

when compared to data from hoof-mounted accelerometers (Starke et al., 2012b).

It has also been determined that an inertial sensor-based system can distinguish

between a positive and negative response to diagnostic anaesthesia of the foot

and objectively assess the effect of a positive response on the trot (Maliye et al.,

2013).

Studies have evolved that compare the data reported from inertial measurement

systems to the experience of Veterinary Clinicians. McCracken et al. (2012)

reported that an inertial sensor system was able to identify lameness in the sole of

the hoof at a lower level of lameness than by the consensus of three experienced

equine Veterinarians. Keegan et al. (2013) also compared data from an inertial

measurement system with subjective lameness examinations from three

experienced equine Veterinarians and reported that an inertial sensor-based

lameness evaluation may enhance but not replace subjective lameness

examination of horses even though all inertial sensor measures for forelimb and

15

hind limb lameness were positively and significantly correlated with results of

subjective evaluations.

It is pertinent to remember that there is now a wide range of inertial measurement

units available on the commercial market but the software may vary tremendously

between systems. A 2016 study sought to quantify differences in two leading

inertial measurement systems and determined that regression-based correction for

systematic differences between the two systems the widths of limits of agreement

values for comparison of straight line trials was within, or only marginally outside

current proposed thresholds of detecting lameness in horses (6 mm for head

movement, 3 mm for pelvic movement) (Pfau et al., 2016a).

1.2.5 Treadmill locomotion

Treadmills have routinely been used in locomotion studies, particularly before the

development of inertial measurement systems that can be used in the field. The

first records of horses on treadmills for research purposes was in 1967 (Persson,

1967). On a treadmill, it is possible to control speed allowing gait to be evaluated

under different conditions, eliminating or controlling for some influencing factors.

However, treadmills have been reported to cause changes in kinematic stride

variables (Fredricson et al., 1983; Leach and Drevemo, 1991; Barrey et al., 1993;

Buchner et al., 1994a; Buchner et al., 1994b). The ability to control speed on a

treadmill may be beneficial, however, it has been reported that speed on a

treadmill is achieved with a higher stride frequency and a longer stride length

(Barrey et al., 1993). There has also been shown to be an increase in stance

16

duration on a treadmill associated with a 9% decrease in speed of the treadmill

belt due to the frictional effect of the vertical force applied through the horse’s limb,

with earlier placement of the forelimbs, greater retraction of both fore and hind

limbs, forelimb preceding hindlimb placement by 22ms and a reduction in vertical

excursions of the hooves and reduced vertical displacement of the hooves

(Buchner et al., 1994a). Clayton and Schamhardt (2007) report that horses use

less energy during treadmill locomotion than during comparative speeds and gaits

overground, which is likely to be due, in part, to an energy transfer from the

treadmill belt to the horse.

It has been documented that although horses adapt very well to treadmill exercise

a period of habituation to the treadmill belt is required before horses move

consistently, with faster gaits habituating quicker and only three five-minute

sessions required to see stabilization of stride kinematics in the trot, whereas the

walk kinematics are not fully stabilized even at the tenth session (Buchner et al.,

1994b). Treadmills have been used countless times in research to study the

movements of the back and limbs, responses to training and in lameness

investigations. Treadmills also now exist that have a force plate embedded in

them which enables consistent measurement of all four limbs over a number of

successive strides (Weishaupt et al., 1996). It is therefore, important to remember

that whilst treadmills can offer the opportunity to collect data from a repeated

number of successive strides at a constant speed, that horses used in treadmill

studies will need a period of acclimatisation to treadmill exercise in each gait.

17

1.3 Hydrotherapy

Equine hydrotherapy is an area of interest that has been rapidly expanding in

recent years. Currently, equine hydrotherapy exists in three main forms:

swimming, cold water static spa therapy, and movement through water in either a

water walker or water treadmill. The water or aqua-treadmill is a piece of

equipment that has rapidly grown in popularity amongst professional and leisure

riders and Veterinary clinicians, and increasingly research is being conducted to

quantify the positive responses reported anecdotally by nearly all that use them.

1.3.1 Human Hydrotherapy

Hydrotherapy is a branch of physiotherapy that has been long recognised for its

benefits which are reported to be relief from pain, swelling and stiffness, joint

mobilisation and increased range of motion, cardiovascular fitness and muscle

strengthening, maintenance & restoration. Historically, ancient human civilisations

are documented to have used hydrotherapy for its physical benefits, notably the

use of hot spas and baths in Roman times, and the Greek philosopher

Hippocrates is reported to have recommended hydrotherapy for relieving medical

conditions (Jackson, 1990). In more recent times, Charles Darwin recognised that

he would not have lived long enough to publish his famous work ‘On the Origin of

Species’ without the hydrotherapy, water cure and homeopathic treatments that he

underwent for his own health, although he was reluctant to admit the benefits of

these alternative therapies for fear of his major works not being taken seriously

(Ullman, 2009). Hydrotherapy specifically involves therapeutic exercise in the

form of swimming or walking (Prankel, 2008). In humans, aquatic exercise has

been used extensively in rehabilitation where there is evidence to suggest that

18

exercise in water is preferable to land exercise for patients suffering from

orthopaedic disease, and that aquatic therapy may aid ease of movement,

swelling reduction and pain relief (Hinman et al., 2007), due to the pressure and

temperature of the water. There are also effects of the water buoyancy and water

resistance (Denning et al., 2012). A recent review article about humans sought to

qualify the differences in physiological and biochemical responses that are related

to aquatic exercise and how these responses differ relative to land-based exercise

so that clinicians can provide more accurate rehabilitation programmes for their

patients finding that pain levels were similar between water calisthenics and land

exercise but that pain levels tended to decrease after underwater treadmill

exercise (Denning et al., 2012).

1.3.2 Animal Hydrotherapy

Specific animal hydrotherapy centres have greatly increased in number due to the

sufficient evidence for the effectiveness of hydrotherapy in humans (McGowan,

2008). Canine hydrotherapy centres in the UK have increased rapidly in the last

decade from zero centres in 1996 to over eighty centres in 2008 (Waining et al.,

2011). The recognition of the importance of rehabilitation after injury has grown

which has resulted in Veterinary Surgeons consulting trained therapists in order to

put together specific rehabilitation programmes for each animal (McGowan, 2008).

They attempt to rehabilitate the animal in a more thorough way so that there is

less chance of a repeat injury or an injury occurring elsewhere before the animal is

fully fit.

Currently, no specific qualification exists for equine hydrotherapy, but there are

regulating bodies for professional practice in animal physiotherapy that strive to

19

promote professional practice in Veterinary physiotherapy to ensure the highest

standards of physiotherapy care are delivered to animals. The main one is the

Association of Chartered Physiotherapists in Animal Therapy (ACPAT) (ACPAT,

2017) which is recognised by BEVA (BEVA, 2017) as a Musculo-Skeletal Allied

Professional. Buchner and Schildboeck stated in 2006 that there were, however,

only a small number of reliable studies offering sufficient evidence for clinical

efficacy of physiotherapy in equines. A level 3 qualification has been accredited in

small animal hydrotherapy, originally pioneered by Hawksmoor in 2002

(Hawksmoor, 2017) but now more widely available throughout the country

including qualifications at Level 4 as well (Moulton College, 2014).

1.3.3 Equine Hydrotherapy

Equine hydrotherapy centres have emerged in a similar way to small animal

therapy centres, as the future of the equine industry has focused far more on

rehabilitation of the sports horse in recent years. Due to the increased focus on

the horse as a performance and sporting animal research has grown in the area of

retraining and rehabilitation from injury. Unlike the racing industry, in sports such

as Eventing, Dressage and Showjumping, the horse’s career as a performance

animal can be increased and prolonged with the use of specific training and

maintenance of the horse’s ability to perform. As a result of this, maintaining

soundness in performance has led to increases in the use of equine physiotherapy

along with other complementary and alternative methods including hydrotherapy.

Due to the lack of clinical evidence many Veterinary Surgeons can be reluctant to

refer an animal for treatment, especially in the current economic climate when

additional therapy would incur further costs to the owner. Although many

insurance companies do have policies that cover alternative therapies, the

20

Veterinary Surgeon can suggest the more traditional recovery approach of trying a

period of rest for the horse which would cost next to nothing, hence current efforts

by some researchers to document findings within the state-of-the-art therapy

centres where they could be invaluable as part of a therapeutic programme and

aid a faster and perhaps better recovery, so by returning an animal to training and

competition in a shorter amount of time.

Much of what is practiced in equine hydrotherapy centres is based on anecdotal

reports and systems of trial and error in conjunction with the Veterinary Surgeon

that first referred the horse. Famously the racehorse Red Rum was trained on

Southport Beach and his successes and recovery from lameness were accredited

to the benefits of bathing in the sea. Currently, there are no laws or legislation that

govern equine hydrotherapy and operation of hydrotherapy equipment and

practices rely on an experienced horse person to safely and confidently train the

horse appropriately in these novel practises.

1.3.4 Modes of equine hydrotherapy

Equine hydrotherapy is limited to three main modes: swimming (whether it be

round pool or straight line but completely non-weight bearing), cryotherapy

(whether it be cold water spa therapy or ice or cold water boots), and water

walking or water treadmill exercise. Specific equine hydrotherapy centres have

been proliferating in the UK over the last twenty years with apparently bigger and

better centres housing more sophisticated and advanced pieces of equipment.

These include the centre at Moulton College in Northamptonshire that uniquely

21

houses a straight-line swimming lane with optional water jets to provide a mild

current for horses to swim against, a cold-water saline spa (CET Equine Spas,

2017) and an aqua-treadmill (FMBs, 2017). This hi-tech equipment is both costly

and requires ample space for safe and effective operation, which seem the most

dominant reasons for opting to have hydrotherapy equipment installed. Centres

such as the one at Moulton College operate commercially so the initial financial

outlay can be considered an investment. However, due to the enormous amount

of positive reports from owners, allied professionals (such as equine

physiotherapists and chiropractors) and Veterinary Surgeons, many private

training and livery yards are opting to include one or more piece of hydrotherapy

equipment for their own use. For example, there are many private equine

swimming pools in racehorse training yards, and many professional event riders

have installed their own cold water spa or at least own a pair of cold water or ice

boots. Aqua-treadmills have increased in popularity enormously over the last few

years, with some equestrian centres specifically opting to invest in this piece of

equipment to use privately and commercially due to the enormous amount of

positive anecdotal evidence and the recent increased surge in scientific evidence

supporting the positive changes the aqua-treadmill can make to a horse’s way of

going (Nankervis et al., 2016; Tranquille et al., 2016; Tabor and Williams, 2017;

Nankervis et al., 2017).

1.3.4.1 Swimming horses

Swimming has been a popular mode of exercising horses for training or as a

means of maintaining cardiorespiratory fitness during layups since the late 1970s

although it is difficult to determine the exercise load. An early study by Thomas et

al. (1980) had the horses dragging weights behind them in an attempt to

22

standardise the exercise load when swimming, but Hobo et al. (1998) determined

the exercise load by evaluating swimming speed, heart rates and blood lactate

concentrations which all indicated that the swimming load had been aerobic.

Bromiley (2000), however, states that swimming exercise is nearly always

anaerobic as the horse is unable to breathe efficiently. Hobo et al. (1998) noted

that arterial haematological parameters degenerated significantly after the

swimming started suggesting that there was an effect of water pressure on the

horse’s body preventing adequate ventilation yet still the exercise was aerobic,

which is contradictory, as if it is aerobic then there is by definition sufficient

ventilation and O2 saturation. Hobo et al. (1998) also interestingly noted that

respiration patterns change from the ‘one-pitch-one-breath’ method that is seen in

field running, suggesting that swimming horses need a greater amount of

ventilation to compensate for the restricted respiratory rates due to the water

pressure on the chest and abdomen, and that expiratory times were longer than

inspiratory times during swimming indicating that longer expiratory times may limit

sudden airway collapse caused by the water pressure preventing a radical

decrease of air space volume and therefore maintain buoyancy.

The physiological responses of humans exercising in water have been well

documented including the energy required for an exercise in water compared to

the same exercise on land, the energy expended, maximal oxygen uptake,

circulation and thermoregulation (Cureton, 1997). However, in horses, studies are

more limited and tend to look at the metabolic and cardiopulmonary responses

(Jones and Hiraga, 2006). The Jones and Hiraga (2006) report made

comparisons to metabolic and cardiopulmonary responses during swimming and

metabolic and cardiopulmonary responses during treadmill galloping and noted

23

some significant conclusions; overall it was found that during swimming horses do

not reach their VO2max and consequently maximum heart rates during swimming

tend to be lower than at VO2max on a dry high-speed treadmill (Jones and Hiraga,

2006).

In general, swimming is considered to have a similar training effect on

cardiorespiratory function but without the load on the legs so is therefore beneficial

for horses with locomotorial disease (Misumi et al., 1994b, 1994c). It is difficult to

draw comparisons between the existing swimming studies as they have all

measured different variables in different ways in both round pools and straight-line

pools. Another study determined that evaluation of the exercise fitness from

swimming horses was comparable to the exercise fitness of running horses and

that poor performance swimming was then a useful indicator to predict poor

performance on the track (Misumi et al., 1994a; 1995). An early study by Irwin

and Howell (1980) stated that when swimming horses they always had a

Veterinary Surgeon present in case of accidents, but there are no legal

requirements in existence today.

1.3.4.2 Cryotherapy

Many livery yards and training yards now own a cryotherapy spa as an added

benefit for their customers and clients as these pieces of equipment are relatively

cheap, easy to install and require less space than a pool or aqua-treadmill. When

studying the literature, although there may be many papers investigating the

efficacy of cryotherapy, particularly in humans, only one paper is directly

concerned with cold spa therapy in horses (Hunt, 2001). Hunt (2001) through

successful trials with twenty-seven horses with various lower leg injuries stated

24

that hypertonic cold water spa therapy was a valuable addition to therapeutic

regimes with horses returning to work sooner and competing successfully without

re-injury. A lot of work has been conducted into the principles of cryotherapy in all

its different forms and recommendations exist for the practical application of the

different forms in different situations particularly with reference to effects of cooling

on survival of tendon cells (Petrov et al., 2003; Reesink et al., 2012). It does not

appear that cold water spa therapy has been included as a mechanism of cooling

by cryotherapy studies and therefore the added benefits of the direct control over

the temperature of the water, the depth of the water and the salinity or mineral

content of the water have not been studied. Humans are less tolerant of cold

water immersion than horses due to the increased blood pressure due to

peripheral vasoconstriction and associated risk of hypothermia (Michlovitz, (1990),

cited in Hunt, (2001)). However, cold water immersion is used regularly in humans

as a popular recovery intervention after exercise although the physiological and

biochemical rationale is still unclear (Bleakley and Davison, 2009).

The composition of natural spa waters has been well documented and research

has been conducted into the therapeutic effects of these waters on various skin

conditions in humans such as psoriasis, and very specific research exists on the

effects of spa waters on cutaneous immunological responses (Joly et al., 2000).

Recently, more investment has been made in the cosmeceutical benefits of spa

waters where anti-inflammatory responses, anti-carcinogenic properties and anti-

radical properties have all been noted (Polefka et al., 2012; Seite, 2013). With

regard to the mechanical effect of balneotherapy (bathing in thermal or mineral

waters) it is reported that immersion allows the mobilisation of joints and the

25

strengthening of muscles with minimal discomfort and reduction in inflammation in

diseases such as rheumatoid arthritis (Nasermoaddeli and Kagamimori, 2005).

Being immersed in water has been documented in humans to bring about

increased diuresis and natriureses, haemodilation, increased cardiac output and

reduced plasma levels (O’Hare et al., 1985) and mechanical stimulation includes a

loss of some body weight thus allowing for easier movement where the body

resists the flotation force and the output and the rhythm of the heart increases and

breathing becomes deeper (Matz et al., 2003). One paper has studied the effect

of immersing horses in warm spring water but this looked at the effect on the

autonomic nervous system and not any mechanical changes finding that

immersion in warm spring water increased parasympathetic nervous activity and

may therefore be a means of relaxation for horses (Kato et al., 2003).

The majority of the research conducted into different forms of balneotherapy has

been conducted in humans and it may not be possible to draw direct comparisons

in horses due to the overt physiological and anatomical differences. However,

work that specifically relates to cryotherapy can probably be extrapolated to

horses in many instances as it seems currently that it is only promoted to immerse

horse lower limbs in cold water where it can be assumed that the make-up of

tendon, ligament and bone structures is comparable to that of humans, but with

there being far less fat and muscle in the distal limb of horses.

1.3.4.3 Aqua-treadmills and walkers

Equine aqua-treadmill exercise has become increasingly popular as a mode of

rehabilitation and training for horses due to the ability to standardise and monitor

many more variables than swimming, thereby potentially being able to deduce the

26

exercise load. Aqua-treadmills in a crude form were first described in the literature

for horses by Auer in 1989, but then no reference appears again until 1999, over

ten years later (Tokuriki et al., 1999). Auer described the idea of the combined

use of a treadmill and a whirlpool whereby jets of water were positioned at

strategic locations so that the horse could work against the resistance of the water

but also then intensifying the massaging action of the water against the limbs and

the effect of the water significantly reducing the shock impact to the legs and

reducing the risk of injury (Auer, 1989). It was suggested at this time that “the

combination of all these features renders such a unit the ideal system for

rehabilitation of a previously injured horse” (Auer, 1989).

There is documentation from the United States Patent Office where an invention

for a piece of equipment for ‘animal exercising, conditioning and therapy’ was

proposed by an E. J. Scanlon in 1969 stating, “This invention is a novel piece of

apparatus wherein the animal is made to run in place on a treadmill, the treadmill

being disposed in an enclosure partially filled with liquid” (Scanlon, 1969). Aqua-

treadmills now feature as the pièce de résistance in equine therapy centres and

training yards. As the technology develops and the price becomes more

affordable, more equestrian enterprises are having them installed. Aqua-

treadmills also exist in a different form of an aqua-walker. These are essentially

horse walkers that are submerged in water, enabling several horses to be

exercised at the same time, thereby giving obvious advantages in time saving but

potentially losing the ability to create individual exercise programmes developed to

the specific needs of each horse. There are fewer aqua-walkers in the UK than in

USA and no research has been detected on the use of an aqua-walker. A further

advantage, however, of an aqua-walker, is that no mechanical belt is needed. It is

27

well documented that equine treadmill locomotion has biomechanical differences

to over ground locomotion (Fredricson et al., 1983; Leach and Drevemo, 1991;

Barrey et al., 1993; Buchner et al., 1994a, 1994b) with a horse apparently first put

on a treadmill for research purposes in 1967 (Persson, 1967). Although it seems

that the aqua-walker may be more beneficial here for research purposes the most

obvious advantage of aqua-treadmills to aqua-walkers is the ability to go in a

controlled straight line for several strides and therefore the aqua-treadmill may

provide locomotion that is more comparable to overground locomotion. Control of

locomotion aided with knowledge of the properties of water such as buoyancy,

viscosity, hydrostatic pressure, surface tension, specific gravity and temperature

effects, should equate to constructive, specific and controlled therapeutic

rehabilitation techniques.

1.4 Current research into aqua-treadmills

As indicated previously, the idea of an aqua-treadmill was first proposed in

academic literature in 1989 as potentially the ideal system to rehabilitate an injured

horse (Auer, 1989) but a patent exists in the USA dating back to 1969 (Scanlon,

1969). The next time aqua-treadmills appear in the literature is not until 1999

where Tokuriki et al. uses an aqua-treadmill as part of a methodology assessing

electromyography activity of different muscles during different types of exercise

suggesting that, at this point, aqua-treadmills are already a recognised and

established exercise and rehabilitation medium for horses. Over the following ten

years (1999-2009) academic literature featuring aqua-treadmills was sparse, but

with the team led by Dr Kathryn Nankervis at Hartpury College made some

significant contributions to introductory aqua-treadmill research. Over the last few

28

years there has been a significant increase in aqua-treadmill research, possibly

due to a combination of the increase in aqua-treadmills available in the UK and

Europe so improving accessibility. But also due to a significant improvement in

technology enabling a wider variety of options for collecting data specifically

relating to gait benefits and adaptations. The aqua-treadmill being an enclosed

metal box, coupled with the ferocious wash of water over the limbs does not lend

itself well to data collection by visual methods or by non-waterproof electronic

devices, making objective quantification of locomotion difficult. Current research

into aqua-treadmills can be divided into a few key areas.

1.4.1 Heart rates

The effects of aqua-treadmill exercise on heart rates has been investigated more

than any other parameter in the literature. A key study from 2006 measured heart

rate responses during acclimation to aqua-treadmill exercise and found that

acclimation to walking on an aqua-treadmill at a depth of mid-radius takes two

fifteen minute sessions with the horses not sedated for heart rates to reach a

steady threshold of around 70 beats per minute (Nankervis and Williams, 2006).

These results confirmed heart rates reported in an earlier study, and in fact the

first study to measure heart rate responses in the literature by Voss et al. in 2002,

where horses were walked or trotted for twenty minutes at water depths of above

carpus, above elbow or on a dry treadmill belt, finding that the most intense

exercise load of trotting for twenty minutes at a water depth of above elbow only

produced a heart rate of 125 beats per minute, and there was only a small

significant difference of 18 beats per minute between the two paces of walk and

29

trot as different workloads, and deducing that aqua-treadmill exercise only

constitutes a medium intensity workload (Voss et al., 2002).

A later study investigated the effect of water temperature on heart rates during

aqua-treadmill exercise and found that increased water temperature from only 13

to 19 degrees Celsius significantly increases heart rate even when exercise

intensity was low (walk) and water depth was considered low at the

scapulohumeral joint (Nankervis et al., 2008a). A further study by the team at

Hartpury College went on to investigate the effect of water depth on heart rates

finding no significant differences in heart rate between each of the wide-ranging

water depths from hoof height (mean HR 62.0 ± 10.2 bpm), to the proximal

interphalangeal joint (mean HR 61.1 ± 8.3 bpm), to carpus (mean HR 60.6 ± 6.7

bpm) and to ulna (mean HR 64.7 ± 8.0 bpm), but this study looked at walk only

(Scott et al., 2010).

A 2003 study reported much higher heart rates of between 120 and 160 beats per

minute when trotting in deep water but with no significant differences found

between water depths ranging from 10 to 80% of the height of the withers or

speeds ranging through walk and trot (Lindner et al., 2003). The higher heart

rates reported here than seen in any of the other studies may therefore be due to

perhaps in increase in water temperature, lack of acclimatisation to the treadmill,

or an increased workload in the trot in deeper water, however, the heart rates are

still not high enough to be consistent with highest maximum values of 210-240

beats per minute described previously in literature (Asheim et al., 1970) therefore

still only producing a medium sized workload which all these studies appear to

30

agree on, possibly due to the published differences in treadmill versus overground

locomotion.

A more recent study of aqua-treadmill exercise and heart rates was in 2011 by the

German team of Lindner et al. (2011) where they again reported slightly different

results with maximum heart rates reaching around 140 beats per minute from an

assortment of variations of depth (ranging from 10 to 77 % of the height of the

withers) and speed (including trotting up to 5.5 metres per second) but that the

fluctuations in heart rate at the different stages of the set exercise tests make it

difficult to use heart rate as an accurate evaluation of conditioning in horses

exercising on an aqua-treadmill and perhaps heart rate results need to be

interpreted differently than from overground studies (Lindner et al., 2011) but it

may well be the temperature of the water that affects heart rates most significantly

(Nankervis et al., 2008a). More recent studies have found similar results where

exercising in an aqua-treadmill does not necessarily produce significant

differences in heart rates (Borgia et al., 2010; Firshman et al., 2015). The

variations in reported heart rates would therefore not appear to be a useful

indicator of workload during equine aqua-treadmill exercise.

1.4.2 Physiological responses

Other physiological responses have often been measured in conjunction with heart

rate studies including blood parameters such as lactate and haemoglobin

concentration. Voss et al.’s (2002) early study measured physiological variables

of lactate and haemoglobin concentration as they have previously been reported

31

as reliable markers to indicate exercise induced changes (Persson, 1969) but

found none of the exercise tests on the aqua-treadmill (walking and trotting and

variable water depths) produced lactacidaemia concluding that horses were able

to provide for the higher oxygen demand via respiration presuming an aerobic

workload. Voss et al. (2002) determined that haemoglobin concentrations

increased with increasing workloads which concurs with well-known previously

reported data that haemoglobin concentration increases as a result of the

mobilization of the erythrocyte reservoir in the spleen during physical exertion

(Persson, 1969).

Lindner et al.’s 2003 and 2011 studies also studied blood lactate concentrations

and found that it was not possible to determine v4 (speed at which a blood lactate

concentration of 4mmol/L is determined under the defined conditions) in horses

exercising on aqua-treadmills as the maximum speed of around 5.5 metres per

second was not fast enough to induce significant responses in blood lactate even

when exercising at the water depth of 80% of withers height did not impose

sufficient additional stress to obtain lactate concentrations of ≥4 mmol/L blood

coming to the same conclusions reported in the heart rate literature, that it is

difficult to evaluate the workload of aqua-treadmill training (Lindner et al., 2003;

Lindner et al., 2011).

A more recent study has assessed not only blood lactate but lactate

dehydrogenase, creatine kinase, aspartate aminotransferase, glucose and

triglyceride levels in horses fed different dietary energy sources and then trained

on an aqua-treadmill and made similar overall conclusions in that the measured

lactate values suggest that the energy requirement of the aqua-treadmill training

32

(even with twenty minutes trotting in water to 85% of the withers) was provided by

an aerobic energy supply as the measured values were below the generally

accepted anaerobic threshold of 2-4 mmol/L but there were more significant

changes in the creatine kinase and aspartate aminotransferase levels which

therefore may be a better indicator of workload intensity (Vincze et al., 2016).

1.4.3 Muscle

An alternative indicator of workload or intensity would be to measure the

electromyographic (EMG) activity of some key muscles during aqua-treadmill

exercise. This was undertaken in the early 1999 study by Tokuriki et al. where

eight skeletal muscles (seven in the forelimb and one in the hindlimb) were tested

during overground walking, swimming and walking and trotting in deep water on

an aqua-treadmill (Tokuriki et al., 1999). Interestingly EMG activity during

swimming tended to have continuous activity within a burst of a cycle but walking

or trotting in the aqua-treadmill often had intermittent activity within a burst and

four horses had more intensified EMG activity at the trot in the aqua-treadmill than

the walk but two horses showed the opposite pattern. The muscle that showed

the most intense EMG activity was the extensor digitorum communis during

walking and trotting in the aqua-treadmill suggesting that this muscle probably

aided in protracting the distal limb against the resistance of the water (Tokuriki et

al., 1999). The water in this study was very deep at 1.2 metres (approximately the

depth of the point of shoulder) so this is very plausible. The study concludes that

trotting in the aqua-treadmill may provide less intensive training for some forelimb

muscles than walking but the depth of the water is important to note here, and

33

frustratingly, the study does not report the results of the EMG activity of the

hindlimb muscle tested.

Over ten years later, another study measured the responses of muscles to aqua-

treadmill measuring muscle lactate, glycogen, and ATP concentrations from

muscle biopsies of the gluteal and superficial digital flexor muscles (SDF) taken

before and after four weeks walking training of up to 20 minutes at a water depth

to the ventral abdomen but found that this exercise protocol was not strenuous

enough to produce any change in the skeletal muscle parameters measured

suggesting a more strenuous exercise protocol was required to therefore have a

training and hypertrophic effect in muscles (Borgia et al., 2010).

A very similar study by the same team in 2015 included modifying the set exercise

test to eight weeks training rising to 40 minutes walking at a depth of the olecranon

but again found no evidence of adaptations within the SDF or gluteal muscles

(Firshman et al., 2015). In 2016, an interesting study was presented at the 8th

International Conference on Canine and Equine Locomotion by a team from

Belgium that measured cross sectional area of fifteen skeletal muscles after eight

weeks of training on an aqua-treadmill for twenty minutes a day in walk at a water

depth of about 40cm. They reported a significant hypertrophy of muscles in the

forelimb, back and hindlimb, particularly muscles involved in elevation and forward

movement of the forelimb, flexion of the hind limb and muscles used for extension

of the spine (Van de Winkel et al., 2016).

34

1.4.4 Gait and locomotory parameters

Recent studies have focussed more on gait and locomotory parameters

suggesting that the development in modern technologies has contributed to this

progression. Gait analysis on an aqua-treadmill was first reported in the literature

in 2010 where the effect of water depth on stride length and frequency was

reported for the first time. The focus was in assessing how many sessions it took

for a steady stride length and frequency to occur in horses that had no experience

of aqua-treadmill exercise (Scott et al., 2010). An accelerometer (Pegasus) was

mounted to the left forelimb on a brushing boot in a waterproof packet and horses

were walked on the aqua-treadmill every day for six days for 15-30 minutes with

water depths increasing to ulna height by the end of the fourth session. Results

showed that horses reached a steady stride frequency and length within the first

six sessions of aqua-treadmill exercise and stride length increased with water

depth up to carpus height but then decreased again when water was increased to

the depth of the ulna and stride frequency decreased at the deeper depths

suggesting that at depths between carpus and ulna the horse may find it easier to

adopt a rounder flight arc by increasing flexion of the hip, stifle and hock joints

(Scott et al., 2010). The Pegasus accelerometer system has previously been

validated against 3D motion capture (Nankervis et al., 2008b) but there does not

appear to be any evidence in the literature of how accelerometers respond once

submerged in water, even if waterproofed, as the forces acting on them are likely

to be altered.

A more recent study investigated flexion and extension of joints of the distal limbs

when walking on an aqua-treadmill at increasing water depths up to stifle level

35

collecting data using a standard video camera at 60Hz finding that any depth of

water above the baseline level of mid hoof significantly increased distal limb joint

flexion, extension and range of motion and as water depth increased percentage

duration of stance phase decreased and swing phase increased in both fore and

hindlimbs (Mendez-Angulo et al., 2013). This supports the earlier work of Scott et

al. (2010) who reported a decreased stride frequency when the water was raised

above carpus level suggesting perhaps that the buoyancy and resistance of the

water therefore slows the passage of the limb. Although two-dimensional

videography is not perhaps the most current and sophisticated gait analysis tool,

the corrections made for the effects of light refraction in water were well

considered here and allayed the potential problems of accelerometer use in water.

Another study compared distal limb range of motion in fore and hindlimbs using

waterproofed accelerometery devices to measure cannon angles from walking on

an aqua-treadmill and walking on a high-speed treadmill finding that walking in

deep water (up to stifle depth) results in a lower forelimb range of motion but a

higher hindlimb range of motion when compared to walking on a dry treadmill

suggesting that there is an effect of drag from the water on the limb (Lefrancois

and Nankervis, 2016). Again, it is important to consider here the application of the

accelerometry devices that have been calibrated for overground use, but the

effects of the system in water has not yet been independently evaluated.

Two recent papers have discussed the effect of aqua-treadmill exercise on the

kinematics of the back, one using high speed video cameras to assess axial

rotation, lateral bending and pelvic rotation at the walk at increasing depths of

water up to shoulder joint level (Mooij et al., 2013) and the most recent using the

36

most sophisticated optical motion capture (Qualisys©) to measure the flexion-

extension range of motion of the thoracolumbar spine and pelvic vertical

displacement in horses walking on an aqua-treadmill at depths of up to the stifle

(Nankervis et al., 2016). Aqua-treadmill training in deep water of elbow and

shoulder depth has been found to significantly increase lateral bending, axial

rotation, and pelvic flexion at each deeper water depth when compared to the

control suggesting at water depths up to the carpus the horse can continue to step

over the water which creates greater axial rotation in the horse’s back and at water

depth to the elbow and shoulder, the horse is forced into a different movement

pattern where the water is too deep to step over so the movement pattern

changed to increase pelvic flexion and reduce lateral bending (Mooij et al., 2013).

Repeated aqua-treadmill training showed no significant changes in movement

patterns in any of the aforementioned parameters (Mooij et al., 2013).

Kinematic data from Nankervis et al. (2016) found that walking in any depth of

water above the control depth of hoof depth was associated with a significantly

greater flexion-extension range of motion in all but the most caudal region (L5) of

the spine which was expected as a result of the increase in stride length with

increasing water depth (Scott et al., 2010). This lack of agreement with the

flexion-extension data at L5 in the 2013 study where walking in all depths of water

was found to increase pelvic flexion may be due to the different data collection

methods used to study this region (Nankervis et al., 2016) with perhaps a bias

towards the data from the 2016 study being more accurate as there is potential

less error assuming correct placement of the recording cameras. Interestingly,

increasing water depth was not found to increase pelvic vertical displacement

(Nankervis et al., 2016) suggesting again, that horses adopt different locomotion

37

patterns to move through the deeper waters. Both these studies only investigated

the changes seen at the walk. There is clear scope to investigate how movement

patterns of the axial skeleton change in the trot using these more sophisticated

high frequency, objective data collection technologies.

1.4.5 Other studies

A handful of other studies have utilised aqua-treadmills for research protocols.

Aqua-treadmills have been proposed as potential useful tools in treating tendon

and ligament injuries mainly due to the beneficial properties of water but no actual

evidence of healing was documented (Adair, 2011). A 2013 study reported that

aqua-therapy benefits osteoarthritis due to its different mechanisms of action in

reducing inflammation and pain and promoting range of motion, but again, no

quantitative data was presented (King et al., 2013a). Later in 2013, the same

team published some quantitative data on the effects of aqua-treadmill training on

postural sway in horses with experimentally induced osteoarthritis, finding that

horses that had undergone a period of aqua-treadmill training walking in shoulder

deep water had improved postural stability in both base-narrow and blindfolded

stance conditions (King et al., 2013b).

Infrared technology was able to non-invasively detect muscle activity and

associated changes in blood flow when horses were exercised on an aqua-

treadmill (Yarnell et al., 2014) but that opens a further array of discussion points

not related to the present study.

38

1.5 Conclusion

The desire to sustain sport horses to remain competitive for longer has resulted in

increased investment in areas of Veterinary Science that promote health,

wellbeing and rehabilitation with more sophisticated pieces of technology being

developed that can accurately evaluate a horse’s ways of going in order to make

major or minor improvements where necessary to give that horse either an extra

competitive edge, or prolonged competitive career.

From substantial anecdotal evidence and an increasing number of clinical trials,

hydrotherapy appears to be a beneficial alternative and/or additional therapy to

promote rehabilitation, healing and training in equines. Despite the increase in

state-of-the-art therapy centres in the UK, evidence of their beneficial use and

application remains predominantly anecdotal but this does not appear to be

deterring Veterinary Surgeons from promoting their use. It is well documented that

movement in water has significant benefits in rehabilitation and the effect of the

temperature of the water has also been studied. The benefits of cryotherapy are

well understood as are the effects of treadmill exercise for fittening and controlled

exercise comparable to that over-ground. Individually, the components of aqua-

treadmill exercise are positive and well understood, but there is little research

evidencing the efficacy of them when they are all added together within the aqua-

treadmill unit. The aqua-treadmill may be a useful tool in manipulating positive

changes to equine locomotion, performance and rehabilitation. There may even

be contraindications that become evident through prolonged use. Therefore, more

research needs to be conducted to be able to quantify the benefits of the aqua-

treadmill so that therapists and Veterinary Surgeons can tailor rehabilitative and

39

training exercise regimes to individuals to faster and better rehabilitate from injury

to improve fitness and prevent the likelihood of reoccurrence.

1.6 Opportunity for Research

Anecdotal reports of the aqua-treadmill include claims that the aqua-treadmill is

useful at rehabilitating horses at any stage of any injury. Such injuries may include

anything from tendon and ligament strains in the lower limbs to overriding dorsal

spinous processes. It has been claimed that if a horse can walk over ground, then

there is no reason why they should not be walking on an aqua-treadmill and

likewise with trot (Baumann, 2015, personal correspondence). To aid in aqua-

treadmill rehabilitation protocols for specific injuries, further research on the effects

of water depth on locomotory parameters is required. It is apparent that there is a

clear gap in the literature to add to the clinical evidence researching the effects of

the aqua-treadmill on equine locomotion to inform and support the use of the

aqua-treadmill in equine rehabilitation and training. In particular, there is an

opportunity for investigation into the effect of water depths on specific locomotory

parameters that are routinely investigated in overground and treadmill studies.

This thesis sets out to make some quantification of the locomotory biomechanics

of the horse, particularly of the axial skeleton during trot as this has not yet been

demonstrated in scientific literature. This study seeks to use the trot as the gait to

study in order to more easily investigate symmetries that may be comparable to

overground studies. And this project sets out to use the most state-of-the-art gait

analysis technologies in order to provide very reputable data for quantitative

40

analysis, that have currently only been used in overground studies and not on an

aqua-treadmill.

This overall research project can be broken down into four distinct aims that

provides the four main chapters of this thesis (Chapters 3, 4, 5 and 6):

• Determine the impact of water depth on vertical pelvis and wither

displacement while trotting on an aqua-treadmill (Chapter 3).

• Determine the effect of side reins on vertical displacements of the pelvis

and withers when trotting on an aqua-treadmill at increasing water depths

(Chapter 4).

• Determine the impact of water depth on mediolateral pelvis and wither

displacements when trotting on an aqua-treadmill (Chapter 5).

• Determine the difference a few centimetres of water makes to pelvis and

wither displacements when trotting on the aqua-treadmill compared to

overground trotting (Chapter 6).

41

CHAPTER 2: Methods

2.0 Methods

Data collection was separated into two distinct data collection methodologies; that

collected by optical motion capture (Qualisys©; ProReflex, Qualisys© Ltd,

Gothenburg, Sweden) between 19th and 21st December 2011 and that collected by

inertial sensor (Xsens©; Xsens© Ltd, Enschede, The Netherlands) between 30th

Oct and 1st November 2013.

2.0.1 Subjects

Overall, 23 horses took part in the trials. Those with incomplete data sets were

removed leaving a total of 18 for analysis to take place (8 for Qualisys© and 10 for

Xsens©). The ultimate reason for withdrawing a horse before data analysis was

due to an incomplete set of data, and the reasons for this included; horses being

withdrawn from the trial due to apparent fatigue, failure of equipment, and a power

cut. One horse took part in both trials (Horse 3 in the Qualisys© trial and Horse 1

in the Xsens© trial) in order to perform a validation between the two systems, so

his data was then discounted from the Xsens© data set to avoid replication. Horse

11 from the Xsens© trial completed all trials apart from side rein data collection on

the aqua-treadmill, so her data is included in most, but not all analyses. All horses

were part of Moulton College working environment where the aqua-treadmill was

located, therefore they all had regular access to the aqua-treadmill prior to the

study and had been on it on multiple occasions but with no real continuity and no

42

specific records kept of the frequency or details of the exercise programmes used.

However, all horses had been exercised on the aqua-treadmill frequently in the

month prior to data collection and were considered habituated. Nankervis and

Williams (2006) identified the necessity of aqua-treadmill habituation and

recommended a minimum of at least two fifteen-minute acclimatising runs to

ensure a steady heart rate.

Breed, size, type, age or fitness of horse were not qualifying factors for this trial.

The majority of horses used were at livery at Moulton College, some belonging to

the college, some on full loan to the college. Two horses taking part in this trial

were not liveried at Moulton College but regular users of the aqua-treadmill. The

horses’ fitness levels were all deemed to be good, as they were all in regular riding

school type work taking part in flatwork, ground schooling and jump sessions for a

maximum three times a day. All horses were considered sound by their owners,

had no history of lameness and had consent from the owner and approval of their

Veterinary Surgeon to take part in the trial. Horses’ heights ranged from 147-

180cm (mean ± 1 SD 165.22 ± 6.18) and ages ranged from 8-19 years (mean ± 1

SD 12.83 ± 3.28).

2.0.2 Protocol

Horses were measured perpendicularly from the ground to four specific locations

on the distal left forelimb which were subsequently used for the different water

depths. These were; the distal phalanx, measured to be half way up the lateral

aspect of the hoof wall (P3); mid metacarpophalangeal joint the centre of the joint

43

located from palpation (mid MCP), the mid-point of metacarpal 3, identified by

locating the centre of both the MCP joint and the carpus joint from palpation and

measuring between them to find the centre (mid MC3); and the mid-point of the

carpus joint located by palpation (mid carpus) (Figure 2.1). Water depth was

matched to the correct depth for each horse during each stage of the trial by use

of a tape measure taped to the outside of the aqua-treadmill window. Water depth

was matched as the horses were moving on the aqua-treadmill during the trials, so

as to not disrupt the horses’ movement patterns and ultimately make the trial last

for a lot longer by having to stop and wait for the water to settle and relying on the

horse to stand still patiently. Horses went on to complete an exercise protocol on

the aqua-treadmill which was a set programme for the Qualisys© trial (Table 2.1)

and a different set programme to include side reins for the Xsens© trial (Table

2.2).

44

Figure 2.1: An illustrative view of the forelimb of the horse identifying the four positions on the limb that were measured to define the four water depths.

4. Mid Carpus – the centre of the carpus joint located by palpation and then measured perpendicularly from the ground.

3. Mid MC3 – the centre of the long cannon bone, meta carpal 3 (MC3) determined from measuring the distance from the centre of the fetlock joint and the carpus joint. This point then measured perpendicularly from the ground.

2. Mid Fetlock – the centre of joint of the fetlock palpated and the distance to the mid-way part of the joint measured perpendicularly from the ground.

1. Mid P3 – the position of the third phalanx within the hoof measured approximately half way up perpendicularly from the ground on the lateral aspect of the hoof wall.

Ground level

3

1

2

4

45

2.1 Use of Qualisys©

2.1.1 Subjects

A total of eight horses were used in this trial. As stated previously, Horse 3 was

used as a test horse for both the Qualisys© and Xsens© trials. These eight

horses’ heights ranged from 157-175cm (mean ± 1 SD 166.25 ± 6.18) and ages

ranged from 11-19 years (mean ± SD 13.13 ± 2.59).

2.1.2 Qualisys© Protocol

Six Qualisys© Oqus 300 cameras (ProReflex, Qualisys Ltd, Gothenburg, Sweden)

were positioned around the aqua-treadmill and the area inside the aqua-treadmill

was calibrated using a standard dynamic wand-based calibration procedure which

defines the orientation of the co-ordinate system with a static frame and wand of

defined length (QTM, 2015). This system is based on passive infrared markers

and infrared cameras. The positive y-axis was orientated in the line of

progression, parallel to the treadmill. The positive z-axis was orientated upward

and the positive x-axis was orientated perpendicular to the y- and z-axes.

A digital video camera (Panasonic HS60 HD) on a tripod was stationed centrally

immediately behind the aqua-treadmill at the height of the tuber sacrale for each

horse when standing as square as possible inside the aqua-treadmill and video

footage was recorded for the whole of each trial.

46

Reflective hemispherical makers were attached to anatomical bony landmarks on

each horse with double sided tape. Markers were attached to the poll or occiput,

tuber sacrale, left and right tuber coxae, left and right ischial tuberosities, withers

(highest dorsally protruding thoracic vertebral process (T4 or T5)), and mid thorax

(approx. half way between T4 and tuber sacrale). Figure 2.2 shows a horse taking

part in the exercise trial on the aqua-treadmill demonstrating the location of these

markers. Horses were taken through an exercise protocol on the aqua-treadmill

(Table 2.1) where data were recorded using the commercially available QTM

software at a frequency of 240Hz.

47

Figure 2.2: A horse on the aqua-treadmill illustrating the location of the light reflective hemi-spherical markers.

Photo credit: Jessica York during data collection, December 2011.

Table 2.1: Aqua-treadmill Qualisys© exercise protocol

WATER DEPTH SPEED TIMING

(minutes) DETAILS

Mid P3 Walk 4.0 Warm horse

Mid P3 Trot 1.5 Collect data

Raise water to mid fetlock Walk 0.5

Mid fetlock Walk 1.5 Collect data

Mid fetlock Trot 1.5 Collect data

Raise water to mid MC3 Walk 0.5

Mid MC3 Walk 1.5 Collect data

Mid MC3 Trot 1.5 Collect data

Raise water to mid carpus Walk 0.5

Mid carpus Walk 1.5 Collect data

Mid carpus Trot 1.5 Collect data

Drain water to mid P3 Walk 4.0 Cool horse

TOTAL TIME 20 minutes

Poll (occiput) Withers (T4 or T5) Mid thorax Tuber sacrale Tuber coxae (left and right) Tuber ischia (left and right)

48

2.2 Use of Xsens©

2.2.1 Subjects

Eleven horses were used in this trial (Horse 1 being the test horse used also in the

Qualisys© trial so his data was discounted for the Xsens© trial). These eleven

horses’ heights ranged from 147-180cm (mean ± SD 164.45 ± 10.22) and ages

ranged from 8-19 years (mean ± SD 12.73 ± 3.69).

2.2.2 Protocol

Horses were fitted with a roller and a snaffle bridle with Newmarket coupling.

Correctly fitted side reins were affixed to the roller to be attached to the bit when

required. All horses used in the trial were familiar with wearing and being

exercised in these basic items of tack. Side reins are a commonly used training

aid for all levels of horses, and all horses in this trial were regularly exercised

overground in them.

The Xsens© inertial sensors used in this study were the wired MT9 model. The

Xsens© sensor units consist of 3-axis accelerometers, 3-axis gyroscopes, and 3-

axis magnetometers that measure 3D linear acceleration, 3D angular velocity, and

3D magnetic field data, measuring 39 x 54 x 28mm, and weighing 35g (Xsens©;

Pfau et al., 2005; Findlow et al., 2008; Valentin et al., 2010). The recorded

accelerations were subsequently double integrated through velocity to determine

displacement of the anatomical bony landmarks in all three axes comparable to

that explained in Pfau et al. (2007).

49

This sensor system was fitted to the horses in turn using Animal Polster©

(Manufactured by Snogg©) and double-sided tape to hold the sensor onto the

Animal Polster©. Large squares of animal polster© were used to securely affix the

sensor to the horses’ coats. The sensors were connected in series to a wireless

transmitter, the XBus (MTx; Xsens, Enschede, The Netherlands), which

synchronised data from all sensors and transmitted it via a wireless digital

telemetry system (Bluetooth) to a nearby laptop (within 50 metres).

Fitting the Xsens© system to the horses consisted of affixing eight inertial sensor

units to the skin overlying bony landmarks commonly used in lameness

investigations: poll, withers (highest dorsally protruding thoracic vertebral process

(T4 or T5), along the thorax (approximately T13) and lumbar (approximately L4)

vertebrae, tuber sacrale, left and right tuber coxae, and top of tail (1st coccygeal

vertebrae). A specially constructed withers mount was used to affix the withers

sensor as per recommendation of the equipment and previous studies (Pfau et al.,

2005). Figure 2.3 shows the ideal location of the sensors denoted by a red dot,

and Figure 2.4 shows the sensors fitted to a horse prior to the start of the trial.

50

Figure 2.3: Visual representation of the placement of the inertial measurement units of Xsens©.

Figure 2.4: A horse fitted with the Xsens© inertial sensors prior to the start of the trial.

Photo credit: Jessica York during data collection, October 2013.

Poll

Withers, approx. T4/T5

Lumbar vertebrae, approx. L4 Tuber sacrale

Thoracic vertebrae, approx. T13

Left and right tuber coxae

Top of tail, approx. 1st coccygeal vertebrae

Poll

Withers, approx. T4/T5

Thoracic vertebrae, approx. T13

Lumbar vertebrae, approx. L4

Tuber sacrale

Left and right tuber coxae

Top of tail, approx. 1st coccygeal vertebrae

51

Horses were trotted in hand on a lunge line (to ensure free and non-restricted

movement by the handler) overground on a flat tarmac road over approximately

50m. The same handler ran each horse. Horses were engaged and encouraged

into a forward-going working trot but not hurried out of their natural rhythm.

Horses were trotted three times, a total distance of approximately 150 metres, to

obtain adequate stride data and to obtain mean values. The first and last few

strides overground were discounted to prevent the effects of acceleration and

deceleration. The laptop and operator were stationed mid-point on the tarmac

route to ensure optimum recording capacity. The road was fenced with moveable

barriers at each end to ensure safety in case the horse became free and to

maintain consistency of trotting distances between trial and horses. The side reins

were then attached and the horses were trotted up three more times in the same

manner. Side reins were fitted according to British Horse Society

recommendations (Auty and Linington-Payne, 2013) at a length so as not to inhibit

or restrict the horse but to work forward into a steady and straight contact.

Walk and trot speeds of each individual horse were also noted immediately from

the GPS and Xsens© software MT Manager, so that the aqua-treadmill speed

could be matched to this overground speed. GPS data loggers have been

reported as accurate methods for determining speed overground in straight lines

(Witte and Wilson, 2004). The Xsens© software collected this information in

metres per second. A simple calculation was used to convert this information into

kilometres per hour in order to match the method used by the aqua-treadmill.

The horses were then moved to the Therapy Centre to complete the aqua-

treadmill exercise protocol straight away (Table 2.2). No rest period was required

52

as the overground trot ups were not deemed to be excessively tiresome. Steps

were placed at the near side of the aqua-treadmill – the opposite side to the

gallery side to assist in easy attachment and detachment of the side reins. The

Xsens© software was used to collect synchronised MTx data via the XBus

collecting data at a frequency of 100Hz. Figure 2.5 shows a horse on the aqua-

treadmill with the sensors attached completing the Xsens© trial. After the protocol

on the aqua-treadmill, horses were cooled down appropriately and had all the

equipment removed, and were put back in their stables safely.

Figure 2.5: A horse on the aqua-treadmill with the Xsens© sensors attached undertaking the set exercise protocol.

Photo credit: Anna Walker during data collection, October 2013.

53

Table 2.2: Aqua-Treadmill Xsens© exercise protocol The exercise protocol used for the horses on the aqua-treadmill for the collection of Xsens© data. Text in bold includes part of the trial where data was collected. ↑ = increase. ↓ = decrease

WATER DEPTH SPEED

(individualised for each horse)

TIMING FREE SIDE REINS XSens©

DATA COLLECT

Raise water Walk 1 min Free - raising water Mid P3 Walk 1 min Free Yes Mid P3 Walk c.10

secs Attach side reins

Mid P3 Walk 1 min Side reins Yes Mid P3 Walk c.20

secs Detach side reins & increase speed

Mid P3 Trot 1 min Free Yes Mid P3 Walk c.30

secs Decrease speed, attach side reins, increase speed

Mid P3 Trot 1 min Side reins Yes ↓speed↑water Walk c.1 min Decrease speed to walk, detach side reins, raise

water.

Mid fetlock Walk 1 min Free Yes Mid fetlock Walk c.10


Mid fetlock Walk 1 min Side reins Yes Mid fetlock Walk c.20


Mid fetlock Trot 1 min Free Yes Mid fetlock Walk c.30


Mid fetlock Trot 1 min Side reins Yes ↓speed ↑water

Walk c.1 min Decrease speed to walk, detach side reins, raise water.

Mid MC3 Walk 1 min Free Yes Mid MC3 Walk c.10


Mid MC3 Walk 1 min Side reins Yes Mid MC3 Walk c.20


Mid MC3 Trot 1 min Free Yes Mid MC3 Walk c.30


Mid MC3 Trot 1 min Side reins Yes ↓speed ↑water

Walk c.1 min Decrease speed to walk, detach side reins, raise water.

Mid carpus Walk 1 min Free Yes Mid carpus Walk c.10


Mid carpus Walk 1 min Side reins Yes Mid carpus Walk c.20


Mid carpus Trot 1 min Free Yes Mid carpus Walk c.30


Mid carpus Trot 1 min Side reins Yes ↓speed Drain Water Cool down

Walk 2 min Decrease speed, detach side reins & empty treadmill. Cool horse down at walk

TOTAL WALK c.16 minutes 40 seconds TOTAL TROT 8 minutes

TOTAL SESSION LENGTH c.24 minutes 40 seconds (25 minutes max.)

54

2.3 Data Processing

2.3.1 QTM (Qualisys©)

Each of the eight markers from each trial was manually tracked and labelled using

the QTM software. Labelled data were then exported to ‘tsv’ (tab separated

values) files to enable data to be imported into MATLAB (R14, The Mathsworks

INC, Natick, MA, US) for further processing and analysis. Firstly, trot data were cut

into individual strides using the minimum sacrum vertical coordinates associated

with mid stance of the left hindlimb. The data at beginning and end of each trial

were excluded to discount for acceleration and deceleration, and only data which

had an apparent clear sine wave were included to discount data where the horse

may have spooked, tripped or lost concentration. This ‘good’ data once cut into

strides, had the minimum and maximum coordinates associated with left and right

trot diagonals extracted in all three axes (vertical, cranio-caudal and medio-

lateral). Only these good data strides were used for further analyses and any

strides that did not correspond with a correct stride pattern were excluded. Every

trial had a minimum of twenty strides used for data analysis.

2.3.2 MTX (Xsens©)

Raw inertial data were calibrated and exported to text files using the MTmanager©

software. Exported data included both accelerations and the Euler angles, namely

roll, pitch and heading. Calibrated acceleration data was processed semi-

automatically into strides using custom written scripts in MATLAB where it was

filtered (using a Butterworth high pass filter with a cut off frequency of 10Hz) and

double integrated through velocity to displacement in all three axes. As with the

QTM data, displacement data were cut into strides, in this case by identifying the

55

vertical acceleration of the left tuber coxae sensor which was used to identify

timings of foot-on of the left hind leg which then allowed automatic detection of the

maximum upward acceleration (approximate time of mid-stance) of the left hind

limb stance phase from the smoothed withers data and the minimum and

maximum vertical values for each stride could be identified and recorded. The

frame numbers for each stride were recorded and utilised to extract the

corresponding Euler angles for calculation of mean pitch and roll amplitudes and

associated comparison between conditions. Every trial had a minimum of twenty

strides used for data analysis.

2.4 Reliability

Reliability for each measure was determined for each condition using the standard

error of the mean (SEM) for each of the trials (c.20 strides per trial). Reliability of

the vertical displacement data is shown in Table 2.3. Reliability of the pitch data is

shown in Table 2.4. Reliability of the mediolateral displacement data is shown in

Table 2.5. Reliability of the roll data is shown in Table 2.6. Reliability of the

absolute position data is shown in Table 2.7. Reliability of the mediolateral flexion

data is shown in Table 2.8. As very low SEMs were calculated for all measures

(range = 0.68 – 2.37 millimetres for displacement data and range = 0.12 – 1.60

degrees for angle data), these data are indicative that the analyses and specific

measures are highly reliable across all conditions. Therefore, differences

calculated between conditions >2.40 mm or >1.60 degrees are likely to be a

consequence of the condition rather than normal variation (error) in the measure.

56

Table 2.3 Vertical Displacement Reliability Data A table of the SEMs for the raw millimetre data of each horse trotting on the aqua-treadmill at a water depth of mid P3 and trotting overground, both with and without side reins, on the left and right parts of the stride for both the pelvis and withers. Considering the low mean SEM for each condition tested here further water depths were not tested.

TROTTING ON THE AQUA-TREADMILL AT A WATER DEPTH OF MID P3 TROTTING OVERGROUND

NO SIDE REINS YES SIDE REINS NO SIDE REINS YES SIDE REINS

PELVIS WITHERS PELVIS WITHERS PELVIS WITHERS PELVIS WITHERS

LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT

QU

ALI

SYS

1 0.96 1.22 0.83 0.88

NO QUALISYS SIDE REIN DATA NO QUALISYS OVERGROUND DATA

2 1.73 1.64 0.94 0.93

3 1.78 1.42 1.63 2.02

4 1.13 1.66 1.03 1.66

5 1.19 1.17 1.01 1.04

6 1.56 1.26 1.43 1.20

7 1.37 1.21 1.16 1.28

XSE

NS

8 1.16 1.70 1.12 1.45 1.31 1.48 0.76 0.95 1.35 1.57 1.58 0.96 1.09 1.52 1.73 0.95

9 1.81 1.61 1.81 1.74 1.47 1.19 2.05 1.64 1.02 1.10 0.88 0.93 1.26 1.05 2.15 1.49

10 1.37 1.34 3.33 3.05 1.26 0.93 1.42 1.37 1.56 1.27 2.19 1.61 1.54 1.10 1.40 2.43

11 1.38 1.20 2.73 2.03 0.95 1.38 2.19 1.20 0.96 0.84 0.64 0.90 0.98 0.71 2.14 1.75

12 1.10 1.04 2.01 1.76 1.06 0.89 1.43 2.13 0.75 0.84 1.66 1.17 0.63 1.16 1.27 1.96

13 0.85 0.64 1.64 1.28 0.95 0.71 1.52 1.07 0.80 1.43 1.16 1.76 0.70 1.03 1.52 1.33

14 1.30 1.15 2.57 1.98 1.44 1.71 1.98 1.23 0.78 0.97 1.40 1.45 0.88 0.74 1.75 1.54

15 1.04 1.34 1.01 0.79 0.87 0.97 1.61 1.88 1.65 1.28 1.66 2.30 0.95 1.71 1.93 0.91

16 1.27 2.08 0.99 1.36 1.10 1.20 2.73 0.78 0.72 0.98 1.01 0.97 0.49 0.65 1.20 1.07

17 1.53 1.93 2.22 2.56 0.65 0.86 3.53 1.21 1.72 1.02 3.32 3.99 1.20 1.15 2.23 1.06

mean SEM 1.32 1.39 1.61 1.59 1.10 1.13 1.92 1.35 1.13 1.13 1.55 1.61 0.97 1.08 1.73 1.45

57

Table 2.4 Pitch Reliability Data A table of the SEMs of the raw pitch data (in degrees) for each horse trotting on the aqua-treadmill at a water depth of mid P3 and trotting overground, both with and without side reins, on the left and right parts of the stride for both the pelvis and withers. Considering the low mean SEM for each condition tested here further water depths were not tested.


NO SIDE REINS YES SIDE REINS NO SIDE REINS YES SIDE REINS

PELVIS WITHERS PELVIS WITHERS PELVIS WITHERS PELVIS WITHERS

LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT

QU

ALI

SYS

1 -

7

NO QUALISYS PITCH DATA NO QUALISYS OVERGROUND DATA

XSE

NS

8 0.61 0.63 2.47 2.44 0.59 0.46 2.10 2.02 0.05 0.07 0.10 0.11 0.08 0.10 0.11 0.15

9 0.97 1.42 0.95 1.06 0.81 1.24 0.90 0.74 0.11 0.16 0.08 0.11 0.14 0.21 0.11 0.12

10 1.23 1.16 1.52 0.99 0.92 1.04 1.43 1.44 0.24 0.17 0.13 0.15 0.33 0.27 0.12 0.17

11 1.54 1.19 1.44 1.37 0.79 1.26 1.07 1.04 0.22 0.22 0.09 0.16 0.24 0.25 0.14 0.12

12 0.89 1.16 1.81 2.61 0.89 1.27 2.56 2.28 0.08 0.11 0.13 0.12 0.12 0.17 0.16 0.16

13 0.68 0.73 1.39 1.60 0.54 0.47 0.94 0.93 0.09 0.10 0.15 0.15 0.10 0.11 0.13 0.12

14 1.47 1.26 1.13 1.30 1.54 0.99 1.24 1.10 0.13 0.17 0.12 0.11 0.20 0.16 0.18 0.21

15 1.67 2.00 2.45 2.67 1.06 1.15 1.78 2.18 0.19 0.36 0.16 0.20 0.13 0.30 0.15 0.20

16 0.65 1.11 1.11 0.95 0.67 0.87 1.33 1.15 0.07 0.08 0.08 0.08 0.06 0.07 0.09 0.08

17 1.42 1.18 0.82 1.01 0.56 0.74 0.96 0.88 0.13 0.24 0.15 0.11 0.18 0.35 0.13 0.13

mean SEM 1.11 1.18 1.51 1.60 0.84 0.95 1.43 1.38 0.13 0.17 0.12 0.13 0.16 0.20 0.13 0.15

58

Table 2.5 Mediolateral Reliability Data A table of the SEMs of the raw mediolateral displacement data (in millimetres) for each horse trotting on the aqua-treadmill at a water depth of mid P3 and trotting overground, both with and without side reins, on the left and right parts of the stride for both the pelvis, withers and poll. Considering the low mean SEM for each condition tested here further water depths were not tested.


PELVIS WITHERS POLL PELVIS WITHERS POLL

NO SIDE REINS

YES SIDE REINS

NO SIDE REINS

YES SIDE REINS

NO SIDE REINS

YES SIDE REINS

NO SIDE REINS

YES SIDE REINS

NO SIDE REINS

YES SIDE REINS

NO SIDE REINS

YES SIDE REINS

QU

ALI

SYS

1 -

7

NO QUALISYS MEDIOLATERAL DATA NO QUALISYS MEDIOLATERAL DATA

XSE

NS

8 0.69 0.43 1.28 0.80 1.76 1.09 0.64 0.78 0.89 0.94 1.37 1.73

9 3.38 0.70 1.71 0.73 3.39 1.80 1.10 1.54 0.75 1.10 1.94 1.88

10 1.43 0.86 1.91 1.11 2.30 1.08 0.79 1.13 0.94 1.21 2.01 1.50

11 2.69 1.21 2.00 1.00 3.57 2.46 0.64 1.09 1.01 0.90 0.93 2.50

12 1.44 1.00 2.83 0.94 4.67 2.46 0.73 0.88 0.61 1.05 1.06 1.30

13 0.83 0.48 1.15 1.00 1.50 1.41 1.35 1.15 1.24 0.82 1.21 0.97

14 1.97 1.15 1.10 1.07 1.38 1.51 1.24 0.00 1.10 0.00 1.50 2.36

15 3.57 1.00 2.19 1.00 2.52 1.51 1.26 1.03 0.96 1.43 2.71 2.72

16 1.45 0.87 1.70 0.83 1.30 1.24 0.91 0.73 1.01 0.71 1.29 1.48

17 1.87 1.29 1.40 0.88 1.30 1.62 1.41 1.45 1.38 1.30 1.50 1.71

mean SEM 1.93 0.90 1.73 0.94 2.37 1.62 1.01 0.98 0.99 0.95 1.55 1.81

59

Table 2.6 Roll Reliability Data A table of the SEMs of the raw roll displacement data (in degrees) for each horse trotting on the aqua-treadmill at a water depth of mid P3 and trotting overground, both with and without side reins, on the left and right parts of the stride for both the pelvis, withers and poll. Considering the low mean SEM for each condition tested here further water depths were not tested.


PELVIS WITHERS POLL PELVIS WITHERS POLL

NO SIDE REINS

YES SIDE REINS

NO SIDE REINS

YES SIDE REINS

NO SIDE REINS

YES SIDE REINS

NO SIDE REINS

YES SIDE REINS

NO SIDE REINS

YES SIDE REINS

NO SIDE REINS

YES SIDE REINS

QU

ALI

SYS

1 -

7

NO QUALISYS ROLL DATA NO QUALISYS ROLL DATA

XSE

NS

8 0.24 0.42 0.29 0.43 0.33 0.17 0.20 0.21 0.28 0.30 0.27 0.33

9 0.92 0.18 0.51 0.26 0.92 0.20 0.27 0.29 0.57 0.42 0.33 0.27

10 0.35 0.24 0.60 0.28 0.26 0.18 0.23 0.25 0.24 0.26 0.74 0.42

11 0.43 0.20 0.76 0.39 0.43 0.30 0.26 0.27 0.20 0.58 0.29 0.56

12 0.56 0.24 0.59 0.71 0.94 0.30 0.25 0.28 0.18 0.23 0.17 0.32

13 0.20 0.20 0.30 0.22 0.19 0.21 0.30 0.18 0.39 0.31 0.40 0.17

14 0.22 0.25 0.39 0.36 0.22 0.16 0.26 0.20 0.34 0.22 0.22 0.40

15 0.32 0.19 0.89 0.56 0.26 0.16 0.25 0.19 0.43 0.36 0.56 0.36

16 0.39 0.25 0.49 0.27 0.46 0.32 0.28 0.22 0.37 0.28 0.14 0.23

17 0.35 0.24 0.38 0.35 0.46 0.26 0.23 0.22 0.43 0.26 0.43 0.51

mean SEM 0.40 0.24 0.52 0.38 0.44 0.23 0.25 0.23 0.34 0.32 0.36 0.36

60

Table 2.7 Absolute Position Reliability Data A table of the SEMs of the raw absolute positions of displacement data (in millimetres) for each horse trotting on the aqua-treadmill at increasing water depths for the most minimum and most maximum parts of the stride for both the left and right parts of the stride for both pelvis and withers. Low mean SEM for each condition were established.

TROT MID P3 TROT MID FETLOCK TROT MID MC3 TROT MID CARPUS

min1 min2 max1 max2 min1 min2 max1 max2 min1 min2 max1 max2 min1 min2 max1 max2

QU

ALI

SYS

1

PEL

VIS

0.66 0.96 0.85 0.91 0.70 0.66 0.70 0.66 1.24 1.16 1.03 0.99 1.15 1.22 1.14 1.02

2 1.71 1.36 1.02 0.73 1.05 1.15 0.99 0.98 1.49 1.38 1.04 1.10 1.46 1.24 1.30 1.47

3 1.12 0.94 0.96 0.87 2.18 1.98 1.63 1.34 2.76 1.99 1.41 1.19 1.98 1.10 1.09 1.14

4 0.85 0.92 1.44 0.90 1.77 1.59 2.29 2.51 4.40 1.71 2.03 2.32 2.19 1.49 2.68 3.88

5 0.97 1.24 0.75 0.85 0.98 1.77 1.03 1.21 1.66 1.56 1.31 1.15 3.73 1.20 1.40 2.11

6 1.00 0.80 0.78 0.82 1.86 1.20 1.06 1.22 1.96 0.76 1.02 0.91 0.83 0.70 0.88 0.81

7 0.97 1.49 0.94 0.97 2.36 2.15 1.75 1.53 1.50 2.35 1.17 1.55 2.82 2.74 1.88 2.30

8 1.08 0.82 0.90 0.87 1.08 1.14 1.08 1.14 1.54 1.75 1.21 1.39 1.94 1.40 1.14 1.12

mean SEM 1.05 1.07 0.95 0.86 1.50 1.45 1.32 1.32 2.07 1.58 1.28 1.32 2.01 1.39 1.44 1.73

QU

ALI

SYS

1

WIT

HER

S

0.45 0.41 0.63 0.81 0.47 0.38 0.60 0.61 0.53 0.50 0.72 0.73 0.35 0.49 0.54 0.56

2 0.60 0.52 0.66 0.68 0.52 0.48 0.55 0.51 0.75 0.90 0.68 0.63 0.53 0.93 0.80 0.66

3 0.78 0.86 0.68 0.79 0.81 0.71 0.93 0.94 0.72 0.62 0.78 0.91 0.56 0.52 0.73 0.60

4 0.83 1.19 1.04 1.55 1.28 1.30 2.17 1.74 1.63 1.35 2.17 3.34 1.39 1.85 4.35 3.59

5 0.57 0.89 0.85 1.20 1.22 1.47 1.38 1.44 1.09 1.54 1.51 1.71 0.65 0.69 1.19 1.01

6 0.60 0.59 0.73 0.68 0.72 0.65 1.15 0.81 0.73 0.65 0.79 0.71 0.49 0.59 0.95 0.91

7 0.69 0.59 1.09 1.02 0.62 0.73 0.84 1.01 0.71 0.60 0.84 0.97 0.75 0.86 1.49 1.32

8 0.91 0.88 0.96 1.03 0.99 1.07 0.99 0.64 1.44 1.27 1.19 1.26 1.27 1.11 1.19 1.12

mean SEM 0.68 0.74 0.83 0.97 0.83 0.85 1.08 0.96 0.95 0.93 1.09 1.28 0.75 0.88 1.41 1.22

61

Table 2.8 Mediolateral Flexions Reliability Data A table of the SEMs of the raw mediolateral flexion data (in degrees) for each horse trotting on the aqua-treadmill at increasing water depths. Low mean SEM for each condition were established.

Mid P3

Mid Fetlock Mid MC3

Mid Carpus

QU

ALI

SYS

3 0.19 0.22 0.21 0.21

4 0.47 0.45 0.54 0.70

5 0.34 0.30 0.28 0.24

6 0.31 0.27 0.19 0.21

7 0.17 0.21 1.93 0.24

8 0.24 0.36 0.48 0.48

mean SEM 0.29 0.30 0.61 0.35

62

2.5 Validating Xsens© against Qualisys©

Optical Motion Capture has long been recognised as the gold standard to assess

3-dimensional movement and orientation with a high degree of accuracy, of which

Qualisys© is one available system. Inertial measurement systems have been

developed as a more practical but objective method to assess movement of which

the Xsens© system is now considered as the gold standard in inertial

measurement technologies (Cutty et al., 2008; Chung et al., 2011; Mourcou et al.,

2015). Xsens© has also been validated against electromagnetic 3D motion

tracking technologies (Saber-Sheikh et al., 2010) and against Optotrack motion

capture technologies (Robert-Lachaine et al., 2017).

Specifically, in horses, Xsens© has been validated against Qualisys© in a study

assessing kinematics of the equine spine where the reflective hemispherical

markers of the Qualisys© system were directly secured to the six inertial

measurement units of the Xsens© system that were located at T6, T10, T13, L1,

S3 (identified by palpation of the respective dorsal spinous processes) plus the left

tuber coxae, and horses were trotted along an 18-metre track surrounded by ten

Qualisys© cameras (Warner et al., 2010). Using this technique, differences

between the two systems were reported to be 8–16 mm (± 2 s.d.) for dorsoventral

movement and 7–11 mm (± 2 s.d.) for mediolateral movement which were both

similar to values reported during treadmill exercise (Pfau et al., 2005). These

values expressed as a percentage were reported as 9–21% accurate for

dorsoventral movement and 10–24% accurate for mediolateral movement, and the

higher mediolateral values were explained by the lower range of movement in this

plane (Warner et al., 2010). Considering that it has previously been reported that

63

even experienced clinicians are not able to reliably detect movement asymmetries

below a level of 25% (Parkes et al., 2009) it was concluded that inertial sensors

provided consistent data between trials and between strides that was sufficiently

comparable to that recorded by motion capture, although due to the relatively

small range of motion in the mediolateral direction, subtle mediolateral

asymmetries might potentially go undetected.

A previous study compared Xsens© to Qualisys© on a treadmill in a similar

fashion focussing on movement at the withers (Pfau et al., 2005) A wand with

three orthogonal arms each bearing a retro-reflective spherical motion capture

marker (the positive x-axis pointing towards the front of the treadmill, the positive

y-axis pointing towards the left side of the treadmill and the positive z-axis pointing

upwards) was fixed to an inertial measurement unit (IMU) located at the withers of

a horse on a treadmill thus rendering a virtual marker at the site of the IMU. Mean

x, y and z displacement and roll, pitch and heading traces obtained using the

inertial sensor were virtually indistinguishable from those obtained using optical

motion capture in walk, trot and canter (Pfau et al., 2005). During trot, compared

with optical motion capture, 50% of the values for the x, y and z displacement

obtained from the inertial sensor were found within (–2.8, +1.4) mm, (–0.9, +0.9)

mm, and (–4.3, +4.9) mm, which compared with the true values derived from the

optical motion capture correspond to a relative error of (–6.5, +3.2)%, (–2.6,

+2.6)%, (–5.6, +6.4)% concluding that the inertial sensors captured cyclical

movements with comparable accuracy to optical motion capture systems (Pfau et

al., 2005).

64

A more recent study investigated dorsoventral flexion-extension of a single horse’s

back when trotting and cantering on a treadmill with synchronised motion capture

and inertial measurement systems (Martin et al., 2014). The difference

(mean ± SD) between the IMU and motion capture during trot were, respectively

for the thoracic and thoracolumbar angles, 0.57 ± 0.44 and 0.65 ± 0.47 degrees

concluding that angle values calculated with IMU data showed acceptable

accuracy consistency for quantification of flexion‐extension movement in a horse's

back (Martin et al., 2014).

Despite current evidence in the literature, it was deemed pertinent to conduct a

validation of the equipment specific to the unique environment of this project and

as such, one horse was used in both the Qualisys© and Xsens© data collection

trials to ensure validity between both operating systems. Qualisys© has been

previously validated in a treadmill environment (Pfau et al., 2005) and Xsens© has

also been validated against Qualisys© in a treadmill environment (Pfau et al.,

2005), but there are currently no reports in the literature that validate either system

in the environment of an aqua-treadmill where the increased mass of metal may

have an effect on the operation of the systems. It is more likely that the

magnetometers of the Xsens© system may be affected by any increased

magnetism of the environment. Ideally, the validation would have occurred

utilising both systems simultaneously, but unfortunately the equipment was only

available for this project separately, two years apart. Also, ideally, both systems

would have been validated simultaneously overground, before moving on to the

aqua-treadmill, but again, the gap of two years in availability of the equipment did

not make this possible and the Qualisys© system was not used at all in

overground trials due to the limited time-frame the equipment was available to us.

65

Due to the cumbersome nature of the Qualisys© equipment, it is best suited to lab

based studies thus expecting that Qualisys© would work well for the aqua-

treadmill in this situation as the aqua-treadmill was enclosed in a barn where

conditions were consistent and the cameras were less likely to be affected by light

anomalies and weather conditions. One horse and one water depth were selected

for a validation trial.

A statistical validation was conducted on one horse that was used in both the

Qualisys© trial and then the Xsens© trial. As discussed, no validation could be

completed for the overground data as the Qualisys© system was not utilised in this

project overground, but both experimental protocols involved trotting on the aqua-

treadmill at a water depth of mid P3. The mean (±SEM) vertical displacements of

the pelvis and withers (in millimetres) of the test horse when trotting on the aqua-

treadmill at a water depth of mid P3 for both Qualisys© and Xsens© are shown in

Table 2.9 and plotted in Figure 2.6.

Table 2.9 Reliability and validation of Xsens© against Qualisys© using one test horse The mean and SEM of the vertical displacements of the pelvis and withers (in millimetres) of the test horse when trotting on the aqua-treadmill at a water depth of mid P3 for both Qualisys© and Xsens©.

QUALISYS XSENS

mean SEM mean SEM

PELVIS Left 105.89 1.28 105.59 1.16

Right 98.97 1.24 100.48 1.70

WITHERS Left 61.88 0.98 69.90 1.12

Right 65.58 0.99 67.72 1.45

Results of a two-way repeated measures ANOVA of the main effect of data

collection system showed no significant difference in the actual values of vertical

displacement amplitude (mm) between Qualisys© and Xsens© (F(1,19) = 4.15, p

66

>0.05) (SEM 1.31). In the pelvis there is a similar displacement difference

between left and right in both Qualisys© and Xsens©. Mean (±SEM) displacement

values for the pelvis on the left using Qualisys© were 105.89 (±1.28) mm

compared to 105.59 (±1.16) mm using Xsens©, a difference of 0.30 mm. Mean

(±SEM) displacement values for the pelvis on the right using Qualisys© were

98.97 (±1.24) mm compared to 100.48 (±1.70) mm using Xsens©, a difference of

1.51 mm. These differences between the two systems (0.3 mm left and 1.51 mm

right) are smaller than previously reported differences by Warner et al. (2010) that

reported differences in dorsoventral movement of 8–16 mm between the two

systems along the back.

In the withers there is also a similar displacement difference between left and right

in both Qualisys© and Xsens©. Mean (±SEM) displacement values for the withers

on the left using Qualisys© were 61.88 (±0.98) mm compared to 69.90 (±1.12)

using Xsens©, a difference of 8.02 mm. Mean (±SEM) displacement values for

the withers on the right using Qualisys© were 65.58 (±0.99) mm compared to

67.72 (±1.45) mm using Xsens©, a difference of 2.14 mm. These differences

between the two systems (8.02 mm left and 2.14 mm right) compare to the

previously reported values of 8-16 mm of difference in dorsoventral movement of

the horse’s back (Warner et al., 2010) although the withers in the current study are

slightly more cranial than the first marker used of T6 in the Warner et al. (2010)

study. The Pfau et al. (2005) study directly recorded vertical displacements of the

withers, but the study did not report mean raw data values, but reported 50% of

the values of dorsoventral displacement of the inertial sensor compared to the

optical motion capture were found within (–4.3, +4.9) mm for trot which suggests a

67

range of 9.2 mm, which again concurs with the differences reported in the present

study.

The difference between left and right changes round in the withers from Qualisys©

to Xsens© and the difference between them becomes smaller. This is perhaps

accounted for by the two years’ difference in data collection and that the horse

may have become more symmetrical in front or perhaps fitter for the second trial.

Two years difference in data collection is a long time, so it is more likely that there

is a difference in the placement of the sensor to the marker used for Qualisys©.

Nevertheless, the overall result shows no significant difference between Qualisys©

and Xsens©. This validation facilitated confidence in the recording systems and

considering the satisfactory levels of agreement reported between the two

systems in this validation trial, it was considered acceptable to amalgamate the

data for the analysis of vertical displacements (seen in Chapter 3).

68

Figure 2.6: Mean (±SEM) vertical displacement amplitude (mm) of the pelvis (left) and withers (right) for the test horse in both the Qualisys© and Xsens© trials at the water depth of mid P3 for both left diagonal pair (green) and right diagonal pair (pink). No significant differences were seen between the two systems.

0

10

20

30

40

50

60

70

80

90

100

110

Qualisys Xsens

Mid P3

Dis

pla

cem

en

t (m

m)

Water Depth and either Qualisys or Xsens

PelvisLeft Right

0

10

20

30

40

50

60

70

80

90

100

110

Qualisys Xsens

Mid P3

Dis

pla

cem

en

t (m

m)

Water Depth and either Qualisys or Xsens

Withers

Left Right

69

2.5 Statistical analyses

A variety of statistical analyses were used throughout the projects and are

described in more detail in each individual chapter. Data analyses were

conducted using IBM SPSS Statistics Version 22 software (IBM Corp. Released

2013). All data were tested for normality throughout all stages of analyses using

Shapiro-Wilk tests and data were normally distributed unless otherwise stated (p >

0.05) (Shapiro and Wilk, 1965). Ultimately, all trials were tested for significance

using repeated measures analysis of variance (ANOVA) tests. Mauchly’s test for

sphericity was investigated and interpreted (Mauchly, 1940) in accordance with the

assumptions of the ANOVA, and where Mauchly's test of sphericity was violated,

the Greenhouse-Geisser correction was applied when interpreting the results

(Laerd Statistics, 2015).

2.6 Health and Safety and Ethics

All horses were fit, well and sound as determined by a Veterinary Surgeon prior to

their involvement in either trial. They also had full owner consent to partake.

Extensive risk assessments were carried out for both data collection trials. For the

Qualisys© trial, risk assessments were carried out to ensure all cables were as

safely covered as possible to enable horses and handlers to negotiate them in a

safe manner with no risk of tripping, damaging the equipment or spooking the

horses. For Xsens© it was necessary to ensure that all wires and cables over the

horses were as safely and securely attached to the horses as possible to ensure

70

that they could not be caught up in anything or hang too low and be tripped over

by the horses or to spook the horses. The laptop receiving the data was

positioned never more than 50 metres away from the horse when data collection

was taking place, and when collecting the data from the aqua-treadmill the laptop

was positioned on a table next to the aqua-treadmill.

71

CHAPTER 3: Vertical displacements of the equine pelvis and withers when trotting on the aqua-treadmill at increasing water depths

3.0 Introduction

Equine aqua-treadmills have increased in popularity as a mode of exercise and

rehabilitation. As discussed in Chapter 1.4 a limited number of scientific studies

have focussed on quantifying the biomechanical movement of the horse during

exercise on the aqua-treadmill. As a result, current practice is based on anecdotal

evidence as little objective, quantitative evidence exists. The study of aqua-

treadmill exercise has been hampered by the ability to collect useful objective,

quantitative data from within the confines of the aqua-treadmill apparatus. Being

an enclosed metal box, access to the limbs is difficult, but the axial skeleton is

exposed leaving scope to collect valid and useful movement data of the horses’

back. The static nature of the treadmill enables a greater quantity of consecutive

stride data to be collected for analysis compared with overground studies. The

aqua-treadmill therefore, should render itself the ideal environment to observe a

horse’s way of going in walk and trot. Utilising sophisticated technologies enables

the consistent observation and analysis of minute changes in movement which

would otherwise be unobservable by the human eye. Comparisons can then be

made to overground movement and movement on a non-water treadmill.

A notable area of interest in movement on an aqua-treadmill is the apparent

vertical lift that is achieved when horses trot through water. Visibly, horses appear

to have to raise their body to aid an increase in movement of the limbs up and

72

over the water but currently there is no evidence to suggest how water depth may

affect this vertical biomechanical movement. This study seeks to determine the

effect of water depth of the vertical displacement amplitude on both the fore

quarters (the withers) and hind quarters (the pelvis) of the horse.

The trot is a symmetrical gait (Hildebrand, 1965) with each of the diagonal limb

pairs (left hind/right fore or right hind/left fore) being dynamically coupled which

results in a symmetrical movement pattern during the left and right diagonal

phases of the stride. Vertical displacement, joint angles and protraction and

retraction within each phase of the trot stride (during left and right diagonal stance

and swing) are symmetrical in a sound horse (Robilliard et al., 2007). Each stride

contains two vertical pelvic displacement minima corresponding to left hind mid-

stance and then right hind mid-stance. The tuber coxae being displaced laterally

from the midline exhibit vertical displacements that differ in amplitude during each

trot diagonal which allows clear identification of left hind (prior to the largest

displacement amplitude) and right hind (prior to the smallest amplitude) stance

phases (May and Wyn-Jones, 1987). Figure 3.1 graphically explains the patterns

seen.

73

Figure 3.1: Graphical representation of the vertical displacements achieved by the os sacrum and the left and right tuber coxae (LTC or RTC) during a single stride from a sound symmetrical horse cut from left hind stance. The vertical arrows indicate amplitude. OS1 and OS2 correspond to the 2 amplitudes per stride of the os sacrum. LC1 and LC2 identify the 2 amplitudes of the left coxae and RC1 and RC2 identify the two amplitudes of the right coxae.

74

Unilateral lameness of horses is characterised by asymmetry in the vertical

displacements between the left and right phases of a stride. The presence of ‘hip

hike’ is often used as a parameter to evaluate lameness in the hind limbs, which is

documented as a rapid upwards movement before foot contact of the lame limb

(May and Wyn-Jones, 1987; Buchner et al., 1993; Pfau et al., 2015).

Displacement asymmetry in horses has reported a symmetry of less than 95% in

the os sacrum to be associated with a lameness score of 1/4 or higher (Peham et

al., 2001, Audigie et al., 2002) with symmetry of less than 85% in the os sacrum

being associated with a lameness score of 2/10 (Church et al., 2009). Even though

these studies use different lameness grading scales, both values are below the

reported 25% threshold which has been reported as the limit for human visual

discrimination performance and the threshold for clinicians to consistently agree

(Parkes et al., 2009).

The vertical displacements of the tuber sacrale are studied in conjunction with the

vertical displacements of the tuber coxae in order to study and assess lameness.

Whilst movement of the sacrum can be used independently to investigate

movement symmetry, the tuber coxae, to date, does not appear to have been

studied in isolation with respect to movement symmetry. One study investigated

mean vertical displacement amplitudes of both parts of the stride in trotters in both

the fore and hindlimbs using an accelerometer on the withers and tuber sacrale,

investigating the influence of speed plus a different track surface (Pauchard et al.,

2014). The consideration of a different type of track surface in trotters may draw

some parallel observations to exercising horses in trot through water. Pauchard et

al. (2014) found that the mean vertical displacement amplitude of the overall stride

(the left and right parts of the stride were not reported independently, only the

75

overall mean) as measured at either the withers or croup in both the fore and hind

limbs, decreased with speed (25-40km/h) regardless of the track type used (either

soft or hard surface); for every 10km/h between 25 and 40km/h the overall vertical

displacement amplitude of the stride decreased approximately 9mm in the withers

and 6mm in the croup. The influence of the hard track was more significant in the

croup than the withers, however, these horses were trotting at speed. This

perhaps suggests that water may provide a cushioning effect equivalent to a softer

track surface that then exhibits vertical displacements not as high as a harder

surface, but a test of speed would also need to be conducted in order to draw true

parallels.

Trotting on an aqua-treadmill does not reach speeds comparable to those

recorded in trotting horses by Pauchard et al. (2014). The aqua-treadmill used for

this study is only calibrated to reach speeds of c.18km/h. Ridden trot speeds

range from 11.5km/h in collected trot, 13km/h in working trot, 16.1km/h in medium

trot and 17.8km/h in extended trot (Clayton, 1994a) while trotting in hand has been

reported to be approximately 14.25km/h (Galisteo et al., 1998). It is previously

reported that the speed at which horses move without being forced is that which

requires the minimum expenditure of energy (Hoyt and Taylor, 1981) which is

arguably the opposite of the requirement of trotting horses where extreme trotting

speeds are reached. Pauchard et al. (2014) concluded that vertical displacements

decrease with speed and the decrease is larger at the withers than the pelvis, also

that a soft track increased the vertical amplitudes of the pelvis compared to a hard

track but not that of the withers. At the highest speeds, track surface no longer

had a significant effect on vertical displacements.

76

To the author’s knowledge, no other studies appear to have investigated the

vertical displacement amplitudes of the pelvis and withers when trotting through

water. This study aims to determine the effect of water depth on the vertical

displacement amplitudes of both the fore quarters (the withers) and hind quarters

(the pelvis) of the horse.

3.1 Aims

The aim of this chapter is to investigate and quantify the effect of water depth on

objective vertical displacements of the horse when exercising in trot on an aqua-

treadmill. Specifically:

1. Analyse the vertical displacement amplitudes of the pelvis and withers when

trotting on an aqua-treadmill at different water depths

2. Investigate the percentage change in vertical displacement amplitudes at the

pelvis and withers when trotting on an aqua-treadmill at different water depths

3. Determine vertical displacement amplitude symmetry at the pelvis and withers

when trotting on an aqua-treadmill at different water depths

4. Investigate the changes in vertical position of the pelvis and withers

throughout a stride cycle when trotting on an aqua-treadmill at different water

depths

77

5. Investigate the percentage change in vertical position of the pelvis and withers

throughout a stride cycle when trotting on an aqua-treadmill at different water

depths

6. Evaluate the pitch amplitudes of the pelvis and withers throughout a stride

cycle when trotting on an aqua-treadmill at increasing water depths.

3.2 Methods and Statistical Analysis

Data were collected and the raw data were processed and analysed as explained

in Chapter 2. Data were processed using custom written scripts in Matlab© and

accelerations were used to calculate displacements before the data were cut into

strides. Only good stride data cut from left hind stance were used for further

analysis. Every trial had a minimum of twenty trot strides used for data analysis.

To compute vertical displacements (namely at the pelvis (tuber sacrale) and

withers (T4-5), a double numerical integration of the accelerations was completed

in conjunction with a high-pass 4th order Butterworth high pass filter applied with a

cut off of 0.5Hz. Vertical displacement as a function of time presents as a double

sinusoidal pattern per stride corresponding to the left and right diagonals (Audigie

et al., 1999; Buchner et al., 1994a; Faber et al., 2001a, 2001b, 2001c; Warner et

al., 2010). All data were cut from the first minima representing mid stance of the

left hind leg determined by the corresponding movement patterns of the markers

or sensors on the sacrum and tuber coxae. Each stride has two minima (min1 and

min2) and two maxima (max1 and max2) indicating the descent and rise for each

diagonal half cycle during trot. Amplitudes were calculated (Equation 1).

78

Equation 1: Ampl1 = Max1 – Min1 and Ampl2 = Max2 – Min2

Mean displacement amplitudes along with their standard deviations and standard

errors were calculated for each trial for each horse.

The inertial measurement units (IMUs) (Xsens©) with their tri-axial

accelerometers, gyroscopes and magnetometers, are able to record synchronised

accelerations, roll, pitch and heading of each sensor on each bony anatomical

landmark. As such, it was possible to cut the data from all sensors in all axes

using the frame numbers recorded from the cutting of the vertical displacement

into cycles. Once cut into strides it is possible to extract the minimum and

maximum values of the pitch angle for each half stride and at increasing water

depths. Pitch amplitudes were then compared between pelvis and withers and

then compared against the displacement amplitudes in order to provide a detailed

analysis of the movement to enable identification of changes in locomotion

associated with different depths of water.

3.2.1 Statistical Analysis

Repeated measures ANOVAS were used for all statistical analyses and these are

described in detail for each individual analysis. Shapiro-Wilk tests for Normality

were used throughout and data were normally distributed unless otherwise stated

(p > 0.05) (Shapiro and Wilk, 1965), and Mauchly’s test for sphericity was

investigated and interpreted (Mauchly, 1940) in accordance with the assumptions

79

of the ANOVA, and where Mauchly's test of sphericity was violated, the

Greenhouse-Geisser correction was applied when interpreting the results (Laerd

Statistics, 2015).

3.3 Results

Data from both the Qualisys© and Xsens© trials were used for this study. Where

data from both trials could be amalgamated (as per Chapter 2, Section 2.5), 17

complete sets of data were available and horses’ heights ranged from 147-180cm

(mean ± SD 165.24 ± 9.11) and ages ranged from 8-19 years (mean ± SD 12.88 ±

3.37). Where analyses utilised only the data from the Qualisys© trial there were 8

complete sets of data available for analysis where horses’ heights ranged from

162-175cm (mean ± SD 165.25 ± 6.18) and ages ranged from 11-19 years (mean

± SD 13.13 ± 2.59). Where analyses utilised only the data from the Xsens© trials

there were 10 complete sets of data available for analysis where horses’ heights

ranged from 147-180cm (mean ± SD 163.20 ± 9.84) and ages ranged from 8-19

years (mean ± SD 12.88 ± 3.84).

3.3.1 Vertical Displacement of the Pelvis and Withers

This analysis utilised the combined data of both the Qualisys© and Xsens© trials.

The vertical displacement values were plotted for the pelvis and withers at each

water depth for both the left and right parts of the stride (Figure 3.2). A repeated

measures ANOVA (2x4x2) was conducted to determine the effects of pelvis or

80

withers, water depth, and the left or right parts of the stride on mean vertical

displacement amplitudes of 17 horses trotting on the aqua-treadmill at increasing

water depths.

There were no interactions between the variables but two significant main effects.

Firstly, there was a large significant effect of the anatomical location (F(1,16) =

42.053, p < 0.001); with post hoc analysis showing that overall, the pelvis had

significantly larger mean vertical displacement amplitudes than the withers with a

mean (±SEM) of 115.28 (±2.52) mm for the pelvis and a mean (±SEM) of 91.53

(±2.36) mm for the withers, a mean (±SEM) significant difference of 23.74 (±2.68)

mm (Table 3.1).

There was also a significant main effect of water depth (F(3,48) = 144.269, p <

0.001), with both the pelvis and withers showing an increase in vertical

displacement amplitudes with each increasing water depth. Table 3.2 shows the

results of the post hoc analysis where there was a significant difference at every

depth.

81

Table 3.1: Post hoc analysis of the significant main effect of pelvis versus withers on vertical displacements of the pelvis and withers when trotting on the aqua-treadmill at increasing water depths. Pairwise comparison with Bonferroni correction. Mean difference shown in mm.

95% Confidence Interval for Differencea

Mean Standard

error Sig.a

Lower Bound

Upper Bound

Pelvis mean estimate 115.28 2.52 110.14 120.41

Withers mean estimate 91.53 2.36 86.74 96.33

Pelvis – Withers mean difference 23.743* 2.678 <0.001 18.294 29.192

Based on estimated marginal means a Bonferroni adjustment for multiple comparisons * The mean difference is significant at the 0.01 level Table 3.2: Post hoc analysis of the significant main effect of water depth on vertical displacements of the pelvis and withers when trotting on the aqua-treadmill at increasing water depths. Pairwise comparison with Bonferroni correction shown. Mean difference shown in millimetres with * denoting significance at the 0.05 level. In the pelvis, means were 96.19, 112.49, 122.29, and 130.14 mm, and in the withers, 74.69, 87.26, 96.81, and 107.37 mm at each increasing depth.


Mean difference

Standard error

Sig.a Lower Bound

Upper Bound

Mid P3

Mid Fetlock -14.43* 1.01 <0.001 -17.26 -11.60

Mid MC3 -24.11* 1.60 <0.001 -28.60 -19.63

Mid Carpus -33.32* 1.57 <0.001 -37.72 -28.91

Mid Fetlock

Mid P3 14.43* 1.01 <0.001 11.60 17.26

Mid MC3 -9.68* 1.08 <0.001 -12.72 -6.64

Mid Carpus -18.88* 1.15 <0.001 -22.11 -15.66

Mid MC3

Mid P3 24.11* 1.60 <0.001 19.63 28.60

Mid Fetlock 9.68* 1.08 <0.001 6.64 12.72

Mid Carpus -9.20* 1.01 <0.001 -12.03 -6.38

Mid Carpus

Mid P3 33.32* 1.57 <0.001 28.91 37.72

Mid Fetlock 18.88* 1.15 <0.001 15.66 22.11

Mid MC3 9.20* 1.01 <0.001 6.38 12.03

Based on estimated marginal means a Bonferroni adjustment for multiple comparisons * The mean difference is significant at the 0.01 level

82

Figure 3.2: Mean (±SEM) vertical displacement (in millimetres) of the equine pelvis (left) and withers (right) at increasing water depths when trotting on an aqua-treadmill at increasing water depths for both left diagonal pair (green) and right diagonal pair (pink) (n=17). Overall the pelvis has significantly larger vertical displacements than the withers (p < 0.01). Vertical displacements increase at each increasing water depth (p < 0.01), a significantly different from all lower water depths.

a

a

a

a

a

a

0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

Mid P3 Mid Fetlock Mid MC3 Mid Carpus

Dis

pla

cem

en

t (m

m)

Water Depth

Pelvis

Left Right

a

a

a

a

a

a

0

10

20

30

40

50

60

70

80

90

100

110

120

130

140


Dis

pla

cem

en

t (m

m)

Water Depth

Withers

Left Right

83

3.3.2 Percentage Change in Displacement of the Pelvis and Withers

It was considered necessary to take the height of the horse into consideration for

analysis as it could be expected that the taller the horse the larger the vertical

displacement amplitude within a stride as all movements are relative. To address

this, a repeated measure ANOVA was conducted using horse height as a co-

variate. The covariate (horse height) met most requirements (correlating with the

dependent variable, not correlating with the independent variable, and having a

linear relationship with the dependent variable), but the analysis failed on

homogeneity of regression with the slope of the regression line for the overall

pelvis scoring -0.745, but for the overall withers data scoring -0.500 where the

slope should have been the same for the different conditions. Therefore, it was

deemed not valid to use horse height as a covariate in the analysis. An alternative

solution was proposed using the proportional increase or percentage change in

vertical displacement amplitudes.

Percentage Change in Displacement

Percentage change in displacement was calculated by using the first water depth

(water at mid P3) as the baseline depth from which to calculate percentage

change for the three increasing water depths. Percentage change in displacement

was calculated for each horse using the mean values that were obtained from the

mid P3 water depth and the means were plotted (Figure 3.3).

To ascertain if this was a suitable method of adjusting for height of horse,

correlations were performed of the percentage data against the millimetre data

and the correlations were analysed. The correlation and confidence intervals were

84

found to be notably smaller in the percentage change data set (mean correlation

coefficient -0.37 for the mm data set versus -0.21 for the percentage data set, a

decrease in the percentage data set of -0.16) suggesting that this adjustment was

a valid procedure to improve the data by removing variability associated with the

height of the horse.

A repeated measures ANOVA (2x3x2) was conducted to determine the effects of

pelvis or withers, water depth, and the left or right parts of the stride on mean

percentage increases in vertical displacement amplitudes of 17 horses trotting on

the aqua-treadmill at increasing water depths.

Results showed a highly significant main effect of water depth (F(2,32) = 79.562, p <

0.001) but also one significant two-way interaction between pelvis or withers and

water depth (F(2,32) = 4.901, p = 0.031). Displacements increased as water depth

increased and post hoc analyses determined a simple main effect of pelvis versus

withers with percentage displacement value scores being mean (±SEM) 10.74

(±4.43) % higher in the withers than the pelvis at the deepest water depth of mid

carpus on the left diagonal pair (p = 0.028) (Table 3.3). Pairwise comparisons of

the simple main effect of water depth showed significant differences at every

combination (p < 0.001), showing significant increases in percentage displacement

at each depth in both the pelvis and withers and on both the left and right diagonal

pairs depth (Table 3.4).

85

Table 3.3: Post hoc analysis of the significant simple main effect of pelvis versus withers on percentage change in vertical displacements of the pelvis and withers when trotting on the aqua-treadmill at increasing water depths. Pairwise comparisons with a Bonferroni correction for multiple comparisons shown. Mean percentage differences shown.


Pelvis to Withers Mean

difference Standard

error Sig.a

Lower Bound

Upper Bound

Mid Fetlock Left -1.82 2.33 0.45 -6.76 3.13

Right 0.91 2.20 0.68 -3.75 5.58

Mid MC3 Left -4.70 2.90 0.12 -10.84 1.44

Right -2.29 2.61 0.39 -7.81 3.24

Mid Carpus Left -10.74* 4.44 0.03 -20.14 -1.33

Right -7.62 4.72 0.13 -17.62 2.37


86

Figure 3.3: Mean (±SEM) percentage increase in vertical displacements of the equine pelvis (left) and withers (right) when trotting on an aqua-treadmill at increasing water depths for both left diagonal pair (green) and right diagonal pair (pink) (n=17). Significant increases in percentage displacement at each depth in both pelvis and withers on both left and right diagonal pair, a significantly different from all lower depths to p < 0.001. b significant difference between pelvis and withers on the left diagonal pair at depth of mid carpus (p = 0.028).

a

ab

a

a

0

10

20

30

40

50

60

Mid Fetlock Mid MC3 Mid Carpus

Pe

rce

nta

ge c

han

ge (

%)

Water Depth

Pelvis

left right

a

ab

a

a

0

10

20

30

40

50

60


Pe

rce

nta

ge c

han

ge (

%)

Water Depth

Withers

left right

87

Table 3.4: Post hoc analysis of the significant simple main effect of water depth on percentage change in vertical displacements of the pelvis and withers when trotting on the aqua-treadmill at increasing water depths. Pairwise comparisons with a Bonferroni correction for multiple comparisons shown. Mean percentage differences shown. A significant effect of depth was found in both the pelvis and withers at both the left and right parts of the stride for each combination.


Mean difference

Standard error

Sig.a Lower Bound

Upper Bound

PEL

VIS

LEFT

Mid Fetlock Mid MC3 -10.87* 1.56 <0.001 -15.03 -6.70

Mid Carpus -19.03* 2.21 <0.001 -24.93 -13.14

Mid MC3 Mid Fetlock 10.87* 1.56 <0.001 6.70 15.03

Mid Carpus -8.17* 1.69 0.001 -12.69 -3.64

Mid Carpus Mid Fetlock 19.03* 2.21 <0.001 13.14 24.93

Mid MC3 8.17* 1.69 0.001 3.64 12.69

RIG

HT


Mid Carpus -18.76* 1.93 <0.001 -23.91 -13.61


Mid Carpus -8.65* 1.71 <0.001 -13.23 -4.07


Mid MC3 9.65* 1.71 <0.001 4.69 13.23

WIT

HER

S

LEFT

Mid Fetlock Mid MC3 -13.75* 2.86 0.001 -21.38 -6.12

Mid Carpus -27.95* 3.82 <0.001 -38.16 -17.74

Mid MC3 Mid Fetlock 13.75* 2.86 0.001 6.12 21.38

Mid Carpus -14.20* 3.18 0.001 -22.70 -5.70


Mid MC3 14.20* 3.18 0.001 5.70 22.70

RIG

HT


Mid Carpus -27.29* 2.99 <0.001 -35.30 -19.29


Mid Carpus -13.98* 2.64 <0.001 -21.05 -6.92


Mid MC3 13.98* 2.66 <0.001 6.92 21.05


88

3.3.3 Symmetry

Analysis of both left and right parts of the trot stride independently allowed

investigation into the symmetry of the horses at each water depth and to

investigate if water depth influences the symmetry of the stride. Symmetry indices

were determined by calculating the difference between the left and right parts of

the stride for each depth (Equation 1). Both the millimetre and percentage score

data were investigated in this study and analysed separately, and the means were

plotted (Figure 3.4).

In the millimetre data, a two-way repeated measures ANOVA (2x4) was conducted

to determine the effect of water depth on the symmetry of the stride in both the

pelvis and withers of 17 horses trotting on the aqua-treadmill at increasing water

depths. Difference value scores were not normally distributed as assessed by

Shapiro-Wilk's test of normality (p > 0.05), however, the ANOVA was run without a

transformation of the data as transforming it did not result in a statistically different

result. There was no significant interaction between water depth and pelvis or

withers (F(3,48) = 0.266, p = 0.850). The main effect of water depth on symmetry

was not significant (F(3,48) = 0.335, p = 0.800), and the main effect of pelvis or

withers on symmetry was not significant (F(1,16) = 0.101, p = 0.755).

In the percentage change data, a two-way repeated measures ANOVA (2x3) was

conducted to determine the effect of water depth on the symmetry of the stride in

both the pelvis and withers of 17 horses trotting on the aqua-treadmill at increasing

water depths. Difference value scores were not normally distributed as assessed

by Shapiro-Wilk's test of normality (p > 0.05), however, the ANOVA was run

89

without a transformation of the data as transforming it did not result in a

statistically different result. There was no statistically significant interaction

between water depth and pelvis or withers (F(2,32) = 0.801, p = 0.458). The main

effect of water depth on symmetry was not statistically significant (F(2,32) = 3.114, p

= 0.084). The main effect of pelvis or withers on symmetry was statistically

significant (F(1,16) = 4.957, p = 0.041) and post hoc pairwise comparisons showed

that overall, the withers are 3.73 % less symmetrical than the pelvis (Table 3.5).

Table 3.5: Post hoc analysis of the significant main effect of pelvis versus withers on symmetry of vertical displacements of the pelvis and withers when trotting on the aqua-treadmill at increasing water depths. Pairwise comparisons with a Bonferroni correction for multiple comparisons shown. Mean percentage differences shown with the withers to be overall 3.73% less symmetrical than the pelvis.


Mean difference

Standard error

Sig.a Lower Bound

Upper Bound

Pelvis Withers -3.73* 1.68 0.041 -7.29 -0.18


90

Figure 3.4: Mean (±SEM) difference between the mean vertical displacements for left and right diagonal for the equine pelvis (purple) and withers (yellow) when trotting on an aqua-treadmill at increasing water depths with raw measurement of mm (left) and percentage change in displacement (right) (n = 17). In the percentage change data, overall the withers are significantly less symmetrical than the pelvis, a (p < 0.05).

0

1

2

3

4

5

6

7

8

9

10


Dif

fere

nce

(m

m)

Water Depth

mm

Pelvis Withers

0

1

2

3

4

5

6

7

8

9

10

11

12

13


Pe

rce

nta

ge D

iffe

ren

ce (

%)

Water Depth

%

Pelvis Withers

a

a

91

3.3.4 Changes in Position

The Qualisys© data can provide information for the actual movement patterns in

terms of looking at the markers in space in time, to show not only the vertical

displacement amplitudes, but the minimum and maximum vertical points reached

through the stride cycle for each marker (both pelvis and withers in this study)

which can give a good visual descriptor of the actual movement patterns within a

stride cycle (n = 8).

Figure 3.5 shows the minimum and maximum vertical positions reached

throughout a stride cycle of both the pelvis and withers. For this dataset, separate

analyses were conducted on the minimum values and then the maximum values.

Minimum positions


pelvis or withers, water depth, and the left or right parts of the stride on the

minimum vertical position reached in the stride cycle of 8 horses trotting on an

aqua-treadmill at increasing water depths. There was no significant three-way

interaction between ‘pelvis or withers’, ‘water depth’ or ‘left or right’ (F(3,21) = 0.260,

p = 0.853). There were no significant two-way interactions. There was, however,

a significant main effect of pelvis versus withers (F(1,7) = 42.038, p < 0.001). Post

hoc analyses of pairwise comparisons show that the pelvis overall has a 43.89 ±

6.77 mm lower minimum position than the withers at the lowest point of the stride

(Table 3.6).

92

There was also a significant main effect of water depth (F(3,21) = 72.122, p < 0.001)

where at each increasing water depth the minimum part of the stride was

significantly lower than the minimum of the preceding depth (means (±SEM) of

20.434 (±20.41), 14.31 (±20.90), 10.96 (±20.01), 5.97 (±21.30) mm) (Table 3.7).

Table 3.6: Post hoc analysis of the statistically significant main effect of pelvis versus withers when analysing the mean minimum part of the stride of 8 horses trotting on an aqua-treadmill at increasing water depths. Pairwise comparison with Bonferroni correction shown. Mean difference shown in mm. (means Pelvis -9.028 versus Withers 34.866mm with a mean difference of 43.894 ± 6.770 mm).


Mean difference

Standard error

Sig.a Lower Bound

Upper Bound

Pelvis Withers -43.89* 6.77 <0.001 -59.90 -27.89

Based on estimated marginal means a Bonferroni adjustment for multiple comparisons * The mean difference is significant at the 0.01 level Table 3.7: Post hoc analysis of the statistically significant main effect of water depth when analysing the mean minimum part of the stride of 8 horses trotting on an aqua-treadmill at increasing water depths. Pairwise comparison with Bonferroni correction shown. Mean difference shown in mm.


Mean difference

Standard error

Sig.a Lower Bound

Upper Bound

Mid P3

Mid Fetlock 6.13* 1.00 0.003 2.49 9.76

Mid MC3 9.48* 1.31 0.001 4.71 14.24

Mid Carpus 14.46* 1.15 <0.001 10.28 18.64

Mid Fetlock

Mid P3 -6.13* 1.00 0.003 -9.76 -2.49

Mid MC3 3.35* 0.80 0.024 0.45 6.26

Mid Carpus 8.34* 0.83 <0.001 5.31 11.37

Mid MC3

Mid P3 -9.48* 1.31 0.001 -14.24 -4.71

Mid Fetlock -3.35* 0.98 0.024 -6.23 -0.45

Mid Carpus 4.98* 0.87 0.004 1.82 8.15

Mid Carpus

Mid P3 -14.46* 1.15 <0.001 -18.64 -10.28

Mid Fetlock -8.34* 0.83 <0.001 -11.37 -5.31

Mid MC3 -4.98* 0.87 0.004 -8.15 -1.82


93

Maximum positions


pelvis or withers, water depth, and the left or right parts of the stride on the most

maximum vertical position reached in the stride cycle of 8 horses trotting on an

aqua-treadmill at increasing water depths (Figure 3.5).

There was no significant three-way interaction between ‘pelvis or withers’, ‘water

depth’ or ‘left or right’ (F(3,21) = 2.529, p = 0.085). There were no significant two-

way interactions. There was one significant main effect of water depth (F(3,21) =

28.581, p < 0.001) where post hoc analyses showed significant increases from mid

P3 to mid fetlock (p = 0.002), from mid P3 to mid MC3 (p = 0.003), and from mid

P3 to mid carpus (p = 0.003) (Table 3.8).

Table 3.8: Post hoc analysis of the statistically significant main effect of water depth when analysing the mean maximum part of the stride of 8 horses trotting on an aqua-treadmill at increasing water depths. Pairwise comparison with Bonferroni correction shown. Mean difference shown in mm.


Mean difference

Standard error

Sig.a Lower Bound

Upper Bound

Mid P3

Mid Fetlock -9.17* 1.35 0.002 -14.09 -4.25

Mid MC3 -14.07* 2.37 0.003 -22.68 -5.45

Mid Carpus -18.79* 3.09 0.003 -30.04 -7.47

Mid Fetlock

Mid P3 9.17* 1.35 0.002 4.25 14.09

Mid MC3 -4.90 1.62 0.114 -10.77 0.98

Mid Carpus -9.63* 2.13 0.016 -17.37 -1.88

Mid MC3

Mid P3 14.066* 2.37 0.003 5.45 22.68

Mid Fetlock 4.90 1.62 0.114 -0.98 10.77

Mid Carpus -4.73 1.71 0.167 -10.94 1.49

Mid Carpus

Mid P3 18.79* 3.09 0.003 7.55 30.04

Mid Fetlock 9.63* 2.13 0.016 1.88 17.37

Mid MC3 4.73 1.71 0.167 -1.49 10.94


94

Figure 3.5: Mean (±SEM) position (mm) in the absolute displacement of the equine pelvis (left) and withers (right) from the minimum starting point (at left hind stance) and maximum vertical point of the stride cycle when trotting on an aqua-treadmill at increasing water depths (n = 8). Left minimum vertical position (min left) (blue). Right minimum vertical position (min right) (light blue). Maximum left vertical position (max left) (green). Maximum right vertical position (max right) (light green). Overall, the pelvis has significantly lower minimum positions than the withers (43.89 ± 6.77 mm, p < 0.001). a At increasing water depths each minimum position was significantly lower than the last (p < 0.001). b Maximum positions all significantly different from water depth of mid P3 (p < 0.01).

a a aa a

a

bb b

bb b

-40-30-20-10

0102030405060708090

100110120130140150160170


Dis

pla

cem

en

t (m

m)

Water Depth

Pelvis

Min left Min rightMax left Max right

a aa

a aa

bb

b

bb

b

-40-30-20-10

0102030405060708090

100110120130140150160170


Dis

pla

cem

en

t (m

m)

Water Depth

Withers


95

3.3.5 Percentage Change in Position

Again, considering the influencing factor of horse height on the vertical

displacement amplitudes and therefore also positions, the percentage change in

position was also investigated and analysed using the information obtained from

the Qualisys© data set. The percentage change data is plotted in Figure 3.6. For

this dataset, separate analyses were conducted on the minimum values and then

the maximum values.

Minimum positions



percentage change in minimum position from the baseline water depth of mid P3

of 8 horses trotting on an aqua-treadmill at increasing water depths. There was no

significant three-way interaction between ‘pelvis or withers’, ‘water depth’ or ‘left or

right’ (F(2,14) = 0.105, p = 0.901), and there were no significant two-way

interactions. There was one significant main effect of water depth (F(2,14) = 28.677,

p < 0.001) where post hoc analyses showed a significant decrease in percentage

change in position from mid fetlock to mid MC3 (p = 0.012) and from mid MC3 to

mid carpus (p = 0.020) (Table 3.9).

96

Table 3.9: Post hoc analysis of the significant main effect of water depth when analysing the mean percentage change in the minimum part of the stride of 8 horses trotting on an aqua-treadmill. Pairwise comparison with Bonferroni correction shown. Mean difference shown in %.


Mean difference

Standard error

Sig.a Lower Bound

Upper Bound

Mid Fetlock

Mid MC3 3.35* 0.80 0.012 0.86 5.85

Mid Carpus 8.29* 1.15 0.001 4.70 11.87

Mid MC3

Mid Fetlock -3.35* 0.80 0.012 -5.85 -0.86

Mid Carpus 4.93* 1.30 0.020 0.87 8.99

Mid Carpus

Mid Fetlock -8.29* 1.15 0.001 -11.87 -4.70

Mid MC3 -4.93* 1.30 0.020 -8.99 -0.87


Maximum positions



percentage change in maximum vertical position from the baseline water depth of

mid P3 of 8 horses trotting on an aqua-treadmill at increasing water depths. There

was no statistically significant three-way interaction between ‘pelvis or withers’,

‘water depth’ or ‘left or right’ (F(2,14) = 2.748, p = 0.098). There was one statistically

significant two-way interaction between pelvis or withers and water depth (F(2,14) =

3.789, p = 0.048).

Post hoc analyses determined that in fact there was no simple main effect of pelvis

or withers but the simple main effect of water depth was significant (Table 3.10) at

the withers on the left diagonal pair where the significant difference lay between

97

the depths of mid Fetlock and mid Carpus with a mean (±SEM) increase of 9.19

(±2.36) % (p = 0.018), and on the right diagonal pair where the significant

differences lay between the depths of mid Fetlock and mid Carpus with a mean

increase of 15.15 (±2.34) % (p = 0.002), and between mid MC3 and Mid Carpus

with a mean increase of 8.67 (±1.88) % (p = 0.007).

Table 3.10: Post hoc analysis of the significant simple main effect of water depth when analysing the mean percentage change in the maximum part of the stride of 8 horses trotting on an aqua-treadmill at increasing water depths. Pairwise comparison with Bonferroni correction shown. Mean difference shown in %.


Mean difference

Standard error

Sig.a Lower Bound

Upper Bound

PEL

VIS

Left

Mid Fetlock Mid MC3 -4.73 2.14 0.187 -11.41 1.95

Mid Carpus -7.10 3.13 0.172 -16.89 2.68

Mid MC3 Mid Fetlock 4.73 2.14 0.187 -0.95 11.41

Mid Carpus -2.37 1.83 0.705 -8.08 3.34

Mid Carpus Mid Fetlock 7.10 3.13 0.172 -2.68 16.89

Mid MC3 2.37 1.83 0.705 -3.34 8.08

Rig

ht


Mid Carpus -7.07 2.72 0.106 -15.56 1.42


Mid Carpus -3.08 2.14 0.581 -9.78 3.62

Mid Carpus Mid Fetlock 7.07 2.72 0.106 -1.42 15.56

Mid MC3 3.08 2.14 0.581 -3.62 9.78

WIT

HER

S

Left


Mid Carpus -9.19* 2.36 0.018 -16.55 -1.82


Mid Carpus -4.79 2.27 0.217 -11.89 2.30

Mid Carpus Mid Fetlock 9.19* 2.36 0.018 1.82 16.55

Mid MC3 4.79 2.27 0.217 -2.30 11.89

Rig

ht


Mid Carpus -15.15* 2.64 0.002 -23.41 -6.88


Mid Carpus -8.67* 1.88 0.007 -14.55 -2.78

Mid Carpus Mid Fetlock 15.15* 2.64 0.002 6.88 23.41

Mid MC3 8.67* 1.88 0.007 2.78 14.55


98

Figure 3.6: Mean (±SEM) percentage change in position (%) from the baseline water level of mid P3 of the equine pelvis (left) and withers (right) from the minimum vertical starting point (at left hind stance) and maximum vertical point of the stride cycle when trotting on an aqua-treadmill at increasing water depths (n = 8). Left minimum vertical position (min left) (blue). Right minimum vertical position (min right) (light blue). Maximum left vertical position (max left) (green). Maximum right vertical position (max right) (light green). Percentage change in minimum displacements decreased at each increasing water depth, a significantly different from all lower water depths (p < 0.01). Maximum values showed significance in the withers only, b significant difference between mid fetlock and mid carpus on the left diagonal pair (p < 0.05), c and on the right (p < 0.01). d significant difference between mid MC3 and mid carpus on the right (p < 0.01).

a

aa

a

-30

-20

-10

0

10

20

30


Pe

rce

nta

ge D

isp

lace

me

nt

fro

m P

3

Water Depth

Pelvis


*

**

*

-30

-20

-10

0

10

20

30


Pe

rce

nta

ge D

isp

lace

me

nt

fro

m P

3

Water Depth

Withers

Min left Min right

Max left Max right

b

c

d

99

3.3.6 Pitch

The inertial measurement units of the Xsens© system could record the pitch of the

sensor throughout the trials. This information was extracted for analysis using

methods previously described and the means were plotted (Figure 3.7).


pelvis or withers, water depth, and the left or right parts of the stride on mean pitch

amplitude throughout a stride cycle of 10 horses trotting on the aqua-treadmill at

increasing water depths. There was no significant three-way interaction between

‘pelvis or withers’, ‘water depth’ or ‘left or right’ (F(3,27) = 0.686, p = 0.568). There

were no statistically significant two-way interactions. There were no statistically

significant main effects. Water depth had no effect on pitch amplitudes throughout

a stride cycle, neither was there an effect of pelvis or withers, or the left or right

part of the stride.

100

Figure 3.7: Mean (±SEM) pitch of the pelvis (left) and withers (right) throughout a stride cycle when trotting on an aqua-treadmill at increasing water depths for both left diagonal pair (green) and right diagonal pair (pink) (n = 10). No significant differences were found in pitch angle.

0

1

2

3

4

5

6

7

8


Me

an A

ngl

e

Water Depth

Pelvis

Left Right

0

1

2

3

4

5

6

7

8


Me

an A

ngl

e

Water Depth

Withers

Left Right

101

3.4 Discussion

The aim of this study was to determine if there was an effect of water depth on the

vertical displacement amplitudes of the fore and hind quarters of the horse (pelvis

and withers) when trotting on an aqua-treadmill. Water depth was found to have a

significant effect on vertical displacement amplitudes of the pelvis and withers, as

water depth increased so did the vertical displacement amplitudes. This was

found to be the case both in terms of the actual values in millimetres and when

looking at the proportionate data in terms of percentage change in displacement at

increasing water depths. It was necessary to look at the proportionate data as the

height of the horse was found to have a positive correlation with the amount of

displacement; as you would expect, taller horses had larger displacements than

shorter horses. It was found not valid to use height as a covariate within statistical

analyses; therefore, using the proportionate change in displacements was

proposed and was found to be effective in reducing the effect of height of horse.

Increase in vertical displacement as water depth increases demonstrates that the

horse is perhaps working harder to push themselves up higher out and over the

top of the water, which may link to the Pauchard et al. (2014) study where a softer

surface increased vertical displacement amplitudes in the pelvis. It may be that

the buoyancy of the water provides a cushioning effect akin to a softer surface

even though the treadmill belt itself would not be categorized as a soft surface.

Studies investigating the effect of water depth on heart rate of horses exercising

on an aqua-treadmill determined that water height had no effect on heart rate

(Voss et al., 2002; Nankervis and Williams, 2006; Nankervis et al., 2008a; Scott et

al., 2010), which would suggest actually no greater energy expenditure or effort.

102

However, all but one of these studies investigated only the walk and not trotting

through water. Voss et al. (2002) determined that there were only small significant

differences of 18 beats per minutes of the mean values of the medians between

walking and trotting and that the small difference plus the relatively low levels of

heart rates and blood lactate concentrations near resting values meant that trotting

on a water treadmill at water depths of carpus or elbow was only a medium-sized

workload and an aerobic activity.

An earlier study that investigated electomyographic (EMG) activity of seven

skeletal muscles in the forequarters and one in the hindquarters of horses both

walking and trotting on an aqua-treadmill showed that the intensity of EMG activity

within the extensor digitorum communis (a forelimb muscle) was higher during

walking on an aqua-treadmill than during trot on an aqua-treadmill (Tokuriki et al.,

1999). Unfortunately, the results of the one muscle investigated in the hindlimb,

the vartus lateralis of the quadriceps femoris were not reported (Tokuriki et al.,

1999) so no comparison can be made between the fore and hind quarters of the

horse. It is known, however, that horses exhibit increased elastic storage at faster

speeds (Biewener, 1998) which may account for the lack of notable EMG activity

reported in the hindlimb (Tokuriki et al., 1999) and the non-significant increase in

workload (heart rates or blood lactate concentrations) at either increasing water

depth (Voss et al., 2002; Nankervis and Williams, 2006; Nankervis et al., 2008;

Scott et al., 2010) or increasing speed from walk to trot (Tokuriki et al., 1999; Voss

et al., 2002). With regards to the results reported in the current study, it is likely

that the increased vertical displacements recorded with increasing water depths

are due to elastic energy storage, the buoyancy of the water and an increased

workload, and it is therefore probable that trotting in water provides a greater

103

stimulus to develop more muscle. Heart rates were not measured in this study so

it is impossible to conclude that increased water depth whilst trotting did or did not

have an effect on heart rate or workload.

The pelvis was found to have vertical displacements significantly greater than

those found at the withers. This suggests that the hind quarters of the horse are in

some way working harder than the forequarters. No other study has made a

comparison between the fore and hindquarters in this way. It is likely that the

hindquarters are more engaged than the forequarters due at large to the anatomy

of the hind limb and the reciprocal apparatus where during the stance phase the

extension of the fetlock joint and stance flexion of the stifle, tarsal and coffin joints

illustrate the shock absorption of the hind limb and in the swing phase the

reciprocal apparatus, which forms the coupling mechanism between stifle and

tarsal joint, also influences the fetlock joint because synchronous flexion and

extension between these three joints have been demonstrated (Wyn-Jones, 1988;

Back et al., 1995b). It is also probable that the hindquarters have an increased

engagement by perhaps tucking under to work harder to create the larger

displacements, however, investigation of the pitch showed no corroboration with

this theory – if the pelvis was significantly ‘tucking under’ then it would be

expressed by the pitch amplitude of the IMU on the tuber sacrale. Increasing

water depth was shown to have no effect on pitch amplitudes throughout the stride

cycle and there was no difference in pitch amplitudes between the pelvis and

withers, in fact if anything, the withers appear to exhibit overall larger pitch

amplitudes than the pelvis. It is very likely that the smaller vertical displacements

seen in the withers are due to the forequarters of the horse being able to

compensate the depth of the water by flexing at the carpus. The slightly larger

104

(but actually non-significant) pitch amplitudes seen in the withers are likely due to

the anatomy of the withers and the difficulty in affixing the IMU to this area (Pfau et

al., 2005). One further study that has extensively investigated trotting on a

treadmill, albeit it not an aqua-treadmill, found that vertical displacement of the

trunk of the horse was reduced (Buchner et al., 1994a) when compared to

overground trotting. Perhaps the small amount of water seen in this study induces

the effect of significantly larger vertical displacements.

When investigating the symmetry of the horse at increasing water depths there

was found to be no statistical significance between increasing water depth and

asymmetry. Although it appeared that there may be an increase in symmetry at a

water depth of mid MC3 in the pelvis, this trend was found to be non-significant.

At the greatest water depth of mid carpus, the horse seemed to become more

asymmetrical, particularly at the withers, but again this trend was not found to be

significant. This suggests that this cohort of horses was perhaps struggling with

this deep water exercise and having to adapt their gait to ‘jump’ over the water in

order to cope with the exertion rather than being able to continue to compensate

the water depth by flexing at the carpus.

The overall mean values obtained for the level of asymmetry between left and right

hind stance were found to be quite high, for example in the pelvis with water at mid

P3, there was approximately 6 millimetres difference between left and right hind

stance. Although the horses used in this study were assessed by a Veterinary

Surgeon prior to the study and were deemed to be sound, this degree of

asymmetry agrees with findings in the literature that state that around 40% of

competition and leisure horses in normal work are found to be lame (Greve and

105

Dyson, 2014) and that the traditional visual assessment of lameness according to

specific grading systems (Dyson, 2011) can often have low inter-observer

agreement (Keegan et al., 1998; Keegan et al., 2010), with within-observer

agreement on lameness levels being higher in more senior clinicians than interns

and residents and concluding that increased experience leads to increased

agreement of lameness identification (Keegan et al., 1998). As horses used in the

current study were only observed by the hydrotherapist and owner and had no

history of lameness, it is possible if observed by more experienced Veterinary

Clinicians some may have been identified as clinically asymmetric i.e. lame. It

has also been demonstrated that there is an observer bias when the observer

knows that a limb or area has received a nerve block showing a difference

therefore in the lameness grade given (Arkell et al., 2006). It is also worth noting

that it has been demonstrated that the human eye struggles to perceive

asymmetries of less than 10-20%, with a difference in hip hike or tuber coxae

asymmetry needing to be 25% to be visible making subtle lamenesses difficult to

detect (Parkes et al., 2009) which is perhaps more pertinent to this study as all

horses that took part were riding school type horses in regular work that were

potentially less well scrutinized than a competition type horse. In this study, it is

unclear if an average asymmetry of 6mm between the tuber coxae would be an

asymmetry obvious enough to detect by eye.

When looking at the percentage change data rather than the actual values, a

different trend exhibited in the withers data where there was an increase in

percentage change of asymmetry at increasing water depths suggesting, that the

deeper the water the more asymmetrical the stride becomes in front. This

potentially indicates that horses whilst engaging the hind quarters and

106

compressing the forelimb to create larger vertical lift may compensate on balance

losing the symmetry of the lift in front. As there are no previous studies

investigating symmetry on an aqua-treadmill there is currently no data available for

comparison. Symmetry could be investigated further with a larger number of

horses of perhaps a more precisely equivalent level of fitness (for example all 3-

year-old race horses) to identify if specific or heavier workloads through deeper

water depths induced changes in symmetry; likewise if shallower water depths

have a more positive effect on symmetry. A study to investigate horses’ symmetry

patterns overground that then subsequent completed a period of exercising in an

aqua-treadmill to analyse changes in patterns of symmetry would be particularly

interesting.

Horses started the stride significantly lower in the pelvis than they did in the

withers. This suggests that the hindquarters compress more to produce the larger

vertical displacements seen, which agrees with previous discussion on the

reciprocal apparatus of the hind limb (Wyn-Jones, 1988; Back et al., 1995b). The

withers perhaps, are more stable, and again, able to compensate the depth of the

water by flexing at the carpus. The starting (most minimum) point of the stride was

also significantly lower in both the pelvis and withers as the water got deeper,

suggesting an increase in both hindlimb and forelimb compression to increase

push off to compensate for the increasing depth of the water. It could be

speculated that this increase in limb compression may be linked to an increase in

stance time or increased peak forces as peak vertical displacement does not alter,

and the same (if not more) force needs to be produced for push off which either

requires a longer time to do it with a longer stance time or results in increased

peak forces which could be detrimental for the musculoskeletal system of the

107

horse and unlikely due to the buoyancy of the water working to reduce concussion.

This, however, conflicts with recent evidence where stance times were found to

decrease as water depth increased but the horses in the previous study were

walking not trotting (Mendez-Angulo et al., 2013), however, studies of treadmill

trotting (without water) have reported an increased stance time as the hoof

remains in contact with the treadmill belt for longer in the retraction phase but the

hoof is placed cranially comparatively with overground trot (Buchner et al., 1994a).

It is suggested that perhaps the addition of water also increases the cranial

placement of the hoof as vertical displacements are increased suggesting an

increased range of motion throughout the limb to create the vertical lift which may

therefore correspond to a subsequent increased cranial placement.

In another study, (Scott et al., 2010) horses walking in water at the level of the

carpus and ulna joints (deep to very deep water) had a lower stride frequency and

a higher stride length of forelimbs compared with those walking at shallower water

depths. This makes sense in that the buoyancy of the water increases and slows

the aerial phase of the walk stride. In the present study in the trot stride there is

likely to be both an increase in the stance time and aerial phase as the horse

compresses the limb to increase push off to create the greater vertical

displacement, which is then large enough to overall slow the stride frequency

meaning that the horses are potentially producing fewer strides in the same

amount of time (as they would overground or with no water) but actually working

harder in those fewer strides which would create a larger workload. Workload or

heart rates were not measured in this study.

108

An increase in limb compression with subsequent greater vertical displacement

may also correspond with a greater range of motion (ROM) being produced

throughout the hindlimb (Mendez-Angulo et al., 2013), although ROM was not

directly measured here, it is biologically logical, as the only other way to increase

vertical displacement would be to drastically increase the work and power output

of the muscles which would be very costly and probably not desirable in the

horses utilised in this study.

The maximum vertical point of the stride was shown to significantly increase with

increasing water depth but there was no significant difference in the magnitude

between the pelvis and the withers, concluding that both the fore and hindlimbs of

the horse reached a similar maximum point of the stride. This makes sense;

otherwise the horse would appear to be very lopsided cranio-caudally when

trotting (a range of 96.43–114.30mm in the pelvis, and 114.24–135.23mm in the

withers).

3.5 Conclusion

Vertical displacement of the equine pelvis and withers increases with increasing

water depth when trotting on an aqua-treadmill and there is a greater displacement

in the pelvis than the withers. Minimum and maximum positions of the pelvis and

withers were found to alter with increasing water depth, with minimum values

decreasing significantly indicating an increase in limb compression during stance.

Maximum vertical positions also increased significantly indicating greater

maximum lift out of the water as a result of the increased compression. No

109

significance was found in displacement symmetry for either the pelvis or withers.

And no significance was found with regards to the change in the pitch of the IMUs

throughout a stride cycle with increasing water depths.

There is plenty of scope for further research here to practically apply the

information found. Further studies to make comparisons to overground trotting

and the vertical displacement values here would be relevant, as would a

longitudinal study to compare starting values before an extended period of aqua-

treadmill exercise to perhaps see if repeated use of an aqua-treadmill has an

overall or long lasting effect on vertical displacements. For example, there is

clearly some muscle training involved with producing the greater vertical

displacements, so how long lived is this training and is there an increasing effect

with increasing aqua-treadmill use? For example, do horses reach a threshold of

increased vertical displacement with repeated aqua-treadmill training?

Understanding how a horse moves on an aqua-treadmill is vital for tailoring

specific therapy treatments and exercise programmes in order to most

successfully rehabilitate the horse from injury and prepare the horse for

competition. Investigation to quantify the effects of increasing water depth on

asymmetric horses should be carried out to further inform and support its

application as a tool for rehabilitation. An investigation into the specific structures

involved and the effects of increased compression should be carried out, including

looking further along the axial skeleton and investigating pelvic and thoracic

rotations, perhaps investigating changes in forces through the limbs, stance and

stride times, distal limb joint kinematics, and changes in ranges of motion in both

walk and trot.

110

CHAPTER 4: The Effect of Side Reins on the Vertical Displacements of the Pelvis and Withers when Trotting on an Aqua-Treadmill at Increasing Water Depths

4.0 Introduction

Chapter 3 of this thesis reported a change in the vertical displacement amplitudes

of the pelvis and withers in horses trotting on an aqua-treadmill at increasing water

depths. When exercising horses overground, some form of training aid is normally

always used on the horse in an attempt to positively alter a horse’s way of going.

There are currently no reports in the literature of training aids being used whilst

horses are exercising on an aqua-treadmill, therefore there is an opportunity for

further research to investigate the effect of a commonly used training aid when

exercising horses on an aqua-treadmill. Therefore, the effect of side reins on the

parameters measured in Chapter 3 could be investigated and compared.

Training aids have been recognised and reported to be beneficial for over 300

years yet there is limited evidence to support their application in the scientific

literature (Cottriall et al., 2009). Side reins are a commonly used British Horse

Society (BHS) approved training aid that are routinely used when working horses

from the ground (i.e. non-ridden work) and are first introduced into the BHS’s

syllabus at Stage 2 when exam candidates are expected to be able to competently

lunge a horse using side reins (BHS, 2017). The aqua-treadmill is a therapeutic

exercise medium for horses so it seems a logical progression to include a training

aid whilst the horses are exercising.

111

Historically, there are few studies that report biomechanical changes when using

training aids, but due to technological advances in measuring biomechanical

parameters there is plenty of opportunity for research in this area. With specific

regards to side reins in the literature, only one study could be found that directly

investigated the use of side reins to measure a musculoskeletal parameter, where

it was hypothesized that using side reins when lungeing horses would increase

electromyographic (EMG) activity of the longissimus dorsi muscle (Cottriall et al.,

2009). In fact, it was found that horses walking and trotting with no side reins on

the lunge had a significantly increased EMG activity of the longissimus dorsi when

compared to walking and trotting with side reins, but that always, the part of the

longissimus dorsi on the inside of the circle had a significantly increased EMG

activity when compared to the outside portion of the muscle (Cottriall et al., 2009).

It would be interesting for some biomechanical parameters other than that of the

left forefoot contact with the ground to have been measured in order to quantify if

any biomechanical changes were occurring at the same time, but of course, this

was not the focus of this particular study. Aqua-treadmills work horses in straight

lines so a comparison between muscle activity on a circle and in straight lines

would also be useful.

Biau et al. (2002) used a gait analysis system to determine the effects of three

different types of reins on kinematic variables. All three types of reins; an elastic

band, a Chambon, and a ‘Back Lift’ showed increased forelimb propulsion at the

trot, with Chambons increasing the dorsoventral activity of the hindlimbs at the trot

and hindlimb propulsion at the walk; the Back Lift increased forelimb dorsoventral

activity both at the walk and trot; and the Chambon increased the activity of the

112

hindlimbs. It was concluded that all three rein types have effects on the kinetic

variables of both forelimbs and hindlimbs but with few significant effects on the

hindlimbs, which is possibly more important in training, but with all three reins

increasing forelimb activity, in spite of the different head-neck angles they

produced, concluding that these training aids are probably better at training the

neck muscles (Biau et al., 2002).

A recent comprehensive study utilised sophisticated technologies to assess rein

tensions in horses trotting in-hand wearing three different types of side rein made

from three materials representing different degrees of elasticity: stiff and inelastic,

stiff elastic, and compliant elastic, and each type of rein was tested at three

different lengths; neutral, short (-10cm) and long (+10cm) (Clayton et al., 2011).

The ultimate recommendation of the study was that side reins with an elasticized

component should be used as they reduce maximal tension and loading rates and

horses seemed more willing to seek a contact with the most elastic side rein as

shown by the higher minimal tensions (Clayton et al., 2011). However, this study

measured nothing other than the rein tensions, rate of loading and impulse on the

reins, and the speeds of the horses when trotting, but did not assess the effect of

these different side reins on any musculoskeletal or biomechanical parameter.

Other studies have focussed on rein tensions in the trot in ridden horses

(Singleton, 2001; Clayton et al., 2003; Clayton et al., 2005; Heleski et al., 2009)

where rein tensions were found to have a consistent and repeatable tension

pattern of two peaks in the tension in one trot stride but in the range of 19 – 80N

dependent on type of horse, bit used, and whether on a straight line or circle, but

113

again, did not discuss the effect of side reins on the locomotory patterns of the

horse.

The ‘Pessoa’ is a modern training aid that was invented by the famous

showjumper and horse trainer Nelson Pessoa (GFS, 2017) and although marketed

to be beneficial for training and developing horses, again there is a lack of

evidence in the scientific literature to support these claims. One study sought to

document temporal, linear and angular variables of the working trot when a

Pessoa training aid was used with a comprehensive data collection strategy using

both high speed video to track skin markers and inertial measurement units and a

Global Positioning System (GPS) to measure stride duration, stride length and

speed (Walker et al., 2013). Ultimately it was determined that the Pessoa may

indeed have benefits for general training and rehabilitation as it appeared to

improve gait scores and encouraged the horse to maintain posture and

lumbosacral flexion without apparent concurrent increase in loading of the fore and

hind limbs. The Pessoa Training Aid incorporates a system of ropes with pulleys

and clips that run around the hocks, down the sides of the horse, clip to the bit and

then from the bit back either in-between the horse’s front legs to a D-ring on the

roller, or back to the roller on the outside of the horse. Due to the low hanging

ropes on the Pessoa, it was deemed unsafe by this team of researchers and

equine hydrotherapists to use a Pessoa training aid in the aqua-treadmill as the

ropes may potentially present as a hazard to the horse within the aqua-treadmill

which overground would not present so much of a problem. Due to the small

space inside the aqua-treadmill and the impossibility of a quick way to evacuate

the horse from the Pessoa or the aqua-treadmill, the Pessoa could potentially

pose a risk with the horse becoming entangled in the ropes if the horse performed

114

any irregular or exaggerated gait or any startling behaviour. The use of side reins

in an aqua-treadmill was not considered unsafe as they could be quickly clipped

on and off the bit and clipped back to the roller out of the way with no potential risk

of the horse becoming entangled in them.

To date, there appear to be no reports in the literature that document

biomechanical changes when exercising horses in side reins. And no evidence to

document the use of any training aid when exercising horses in an aqua-treadmill.

For this study, it was hypothesized that the use of side reins would potentially have

an impact on the biomechanical parameters measured previously with changes in

the vertical displacement amplitudes achieved at the pelvis and withers.

4.1 Aims

The aim of this chapter was to investigate and quantify the effect of side reins on

the vertical displacements of the pelvis and withers when trotting on an aqua-

treadmill at increasing water depths. Specifically:

1. Analyse the effect of side reins on the vertical displacement amplitudes of

the pelvis and withers when trotting on an aqua-treadmill at increasing

water depths.

2. Analyse the effect of side reins on the percentage change in vertical

displacement amplitudes of the pelvis and withers when trotting on an

aqua-treadmill at increasing water depths.

115

3. Analyse the effect of side reins on the symmetry of the vertical

displacement amplitudes of the pelvis and withers when trotting on an


4. Analyse the effect of side reins on the pitch amplitudes of the pelvis and

withers throughout a stride cycle when trotting on an aqua-treadmill at

increasing water depths.


Data were collected as detailed in Chapter 2.2 and the raw data were processed

and analysed as explained in Chapter 2.3. However, it was only the data collected

during the Xsens© trial that was used for this study as it was during the initial

Qualisys© data collection trial where it was observed that the use of side reins

may be an interesting and valid addition to the investigation. As previously, data

were processed using custom written scripts in Matlab© and only good stride data

cut from left hind stance were used for further analysis. Every trial had a minimum

of twenty strides used for data analysis. To compute vertical displacements

(namely at the pelvis (tuber sacrale) and withers (T4-5), a double numerical

integration of the accelerations was completed with a high-pass 4th order

Butterworth filter applied with a cut off of 0.5Hz. Methodology procedures have

been reported in Chapter 3.2 so will not be repeated here.

116









Statistics, 2015).

4.3 Results

The vertical displacement amplitudes of the pelvis and withers when trotting on an

aqua-treadmill at increasing water depths without side reins are reported in

Chapter 3.3.1. However, for this analysis, only the horses used in the Xsens©

data collection trial were included as it was only these horses that had the use of

side reins incorporated into their exercise protocol. As such there were a total of

10 good and complete sets of data from the Xsens© data trial to complete this

analysis. Horses heights ranged from 147-180cm (mean ± SD 163.2 ± 9.84) and

ages ranged from 8-19 years (mean ± SD 12.90 ± 3.84).

Considering that only the 10 Xsens© horses were used for this analysis, the

results of vertical displacement of the pelvis and withers may have been slightly

different from the values reported in Chapter 3 which includes the amalgamated

117

results of both the Qualisys© and Xsens© trials. Table 4.1 shows the comparison

of mean values obtained from the Qualisys© and Xsens© data collection trials

together (Q+X, n=17) and then the Xsens© trial on its own (n=10). Results of a

repeated measures ANOVA showed no significant difference (p = 0.456) between

the values with the results from Qualisys© plus Xsens© and the values from just

the Xsens© trial with all results being within 5% for each corresponding value

(Table 4.1). Therefore, it can be deemed valid to use only the results of the

Xsens© trials in further analyses where necessary. A further validation of Xsens©

against Qualisys© was completed prior to data analysis and can be found in

Chapter 2.

Table 4.1: A comparison of the mean vertical displacement values (in mm) for the pelvis and withers obtained from the amalgamation of both Qualisys© (Q) and Xsens© (X) data collection trials (Q+X, n=17) and then just the values obtained from the Xsens© data collection trial (X, n=10). Percentage difference between the two means is shown (%diff).


Q+X X %

diff Q+X X

% diff

Q+X X %

diff Q+X X

% diff

PELVIS Left 95.97 95.77 0.21 112.06 109.31 2.46 122.22 116.99 4.28 129.85 125.46 3.38

Right 96.40 97.40 1.04 112.91 111.65 1.12 122.37 119.26 2.54 130.44 126.15 3.29

WITHERS Left 73.16 71.60 2.13 86.19 84.73 1.70 95.38 93.53 1.94 105.70 106.16 0.43

Right 76.23 76.19 0.06 88.32 88.71 0.44 98.24 97.33 0.93 109.04 108.96 0.08

118

4.3.1 The effect of side reins on vertical displacements of the pelvis and

withers

A four-way repeated measures ANOVA (2x4x2x2) was conducted to determine the

effect of side reins on the vertical displacement of the pelvis and withers of ten

horses trotting on an aqua-treadmill at increasing water depths. Mean results

were plotted (Figure 4.1).

The four-way interaction between pelvis or withers, depth, use of side reins and

left or right was not significant (F(3,27) = 0.154, p = 0.927). There were no

significant three-way or no significant two-way interactions. The main effect of

side reins was not found to be significant (F(1,9) = 3.839, p = 0.082). Otherwise

the same trends, significances and effects that were seen in Chapter 3.3.1 were

also seen here; there was a significant effect of pelvis versus withers (F(1,9) =

47.712, p < 0.001) with the pelvis showing overall greater displacement amplitudes

than the withers, and a significant effect of water depth (F(3,27) = 137.239, p <

0.001), where increasing water depth showed an increase in vertical displacement

amplitudes.

119

Figure 4.1: Mean (±SEM) vertical displacement (in millimetres) of the equine pelvis (left) and withers (right) at when trotting on an aqua-treadmill at increasing water depths both without and with side reins for both left (green) and right (pink) diagonal pair (n=17). No side reins indicated by dots. Side reins indicated by dashes. There was no significant effect of side reins. Overall the pelvis has significantly larger vertical displacements than the withers (p < 0.01). Vertical displacements increase at each increasing water depth (p < 0.01), a significantly different from all lower water depths

aa

aa a a

a aa a a a

0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

NO SR YES SR NO SR YES SR NO SR YES SR NO SR YES SR


Dis

pla

cem

en

t (m

m)

Water Depth

Pelvis

Left Right

a a

a a

a a

a aa

a

a a

0

10

20

30

40

50

60

70

80

90

100

110

120

130

140



Dis

pla

cem

en

t (m

m)

Water Depth

Withers

Left Right

120

4.3.2 Percentage change in vertical displacements of the pelvis and withers

with and without side reins

As explained previously in Chapter 3.3.2, the height of the horse is positively

correlated with vertical displacement amplitudes, so once again, it is worth

considering the proportionate data in terms of percentage changes from the

baseline level of water at mid P3 to investigate any impact of side reins. Figure

4.2 shows the data from the percentage change in displacement from the baseline

level of water at mid P3 to the increasing water depths at both the pelvis and

withers.

A four-way repeated measures ANOVA (2x3x2x2) was conducted. In

concordance with the results Chapter 4.3.1, there was no effect of side reins on

the change in vertical displacement amplitudes of the pelvis and withers at

increasing water depths when trotting on an aqua-treadmill from a baseline level of

water at mid P3 (F(1,9) = 0.218, p = 0.652).

In concordance with the results reported in Chapter 3.3.2 there is a highly

significant effect of increasing water depth increasing the changes in displacement

amplitudes (F(2,18) = 60.982, p < 0.001) and also a significant difference overall

between pelvis and withers (F(1,9) = 10.962, p = 0.009).

121

Figure 4.2: Mean (±SEM) percentage change in vertical displacements of the equine pelvis (left) and withers (right) when trotting on an aqua-treadmill at increasing water depths without and with side reins for both left (green) and right (pink) diagonal pair (n=10). No side reins indicated by dots. Side reins indicated by dashes. There was no significant effect of side reins on the percentage change in vertical displacement amplitudes.

0

10

20

30

40

50

60

70

NO SR YES SR NO SR YES SR NO SR YES SR

Mid Fetlock Mid MC3 Mid Knee

Pe

rce

nta

ge c

han

ge (

%)

Water Depth and Side Rein (SR) status

Pelvis

Left Right

0

10

20

30

40

50

60

70



Pe

rce

nta

ge c

han

ge (

%)


Withers

Left Right

122

4.3.3 The effect of side reins on symmetry of vertical displacement

amplitudes

Chapter 3 (3.3.3) established that there was no effect of water depth on the

symmetry of the horse. This section seeks to quantify the effect of side reins on

symmetry. Both the millimetre and percentage score data were investigated in

this study and analysed separately, and the means were plotted (Figure 4.3).

In the millimetre data, a repeated measures ANOVA (2x4x2) was conducted to

determine the effect of side reins on the symmetry of the stride in both the pelvis

and withers of 10 horses trotting on the aqua-treadmill at increasing water depths.

There was no significant three-way interaction between pelvis or withers, water

depth and side reins (F(3,27) = 1.623, p = 0.229). There were no significant two-

way interactions, and the main effect of side reins on symmetry was not significant

(F(1,9) = 0.285, p = 0.606). As reported in Chapter 3.3.3, the main effect of pelvis

or withers on symmetry was not significant (F(1,9) = 0.648, p = 0.442), and the main

effect of water depth was also not significant (F(3,27) = 1.072, p = 0.377).

In the percentage change data, a repeated measures ANOVA (2x3x2) was

conducted to determine the effect of side reins on the symmetry of the stride in

both the pelvis and withers of 10 horses trotting on the aqua-treadmill at increasing

water depths in terms of the percentage change in displacement. There was no

statistically significant three-way interaction between pelvis or withers, water depth

and side reins (F(2,18) = 2.843, p = 0.085), and there were no statistically significant

two-way interactions.

123

The main effect of side reins on symmetry on the percentage change in

displacement was not statistically significant (F(1,9) = 1.444, p = 0.260). As

reported in Chapter 3 (3.3.3) and again seen here, the main effect of pelvis or

withers on percentage change in displacement symmetry was statistically

significant (F(1,9) = 9.366, p = 0.014), but there was no main effect of water depth

on percentage change symmetry (F(2,18) = 1.758, p = 0.201).

Side reins have no impact on displacement symmetry of horses trotting on an


124

Figure 4.3: Mean (±SEM) difference between the mean vertical displacements following left and right hind stance for the equine pelvis (purple) and withers (yellow) when trotting on an aqua-treadmill at increasing water depths with raw measurement of mm (left) or considering the percentage change in displacement (right) both without AND with side reins (n=10). No side reins indicated by dots. Side reins indicated by dashes. There was no effect of side reins on displacement symmetries either in the raw mm data or in the percentage change data.

0

1

2

3

4

5

6

7

8


Mid P3 Mid Fetlock Mid MC3 Mid Knee

Dif

fere

nce

(m

m)


mm

Pelvis Withers

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16



Dif

fere

nce

(%

)


%

Pelvis Withers

125

4.3.4 Impact of side reins on pitch

The effect of side reins on the pitch of the pelvis and withers throughout a stride

cycle when trotting on an aqua-treadmill at increasing water depths was

investigated and the means plotted (Figure 4.4).

Results of a four-way repeated measures ANOVA (2x4x2x2) showed no significant

four-way interaction (F(3,27) = 1.187, p = 0.333), no significant three-way

interactions and no significant two-way interactions. There was no main effect of

side reins on changes in pitch (F(1,9) = 1.083, p = 0.325).

126

Figure 4.4: Mean (±SEM) pitch of the pelvis (left) and withers (right) throughout a stride cycle when trotting on an aqua-treadmill at increasing water depths both without and with side reins for both left (green) and right (pink) diagonal pair (n = 10). No side reins indicated by dots. Side reins indicated by dashes. There was no effect of side reins on changes in pitch amplitude in the pelvis or withers.

0

1

2

3

4

5

6

7

8



An

gle

(°)


Pelvis

Left Right

0

1

2

3

4

5

6

7

8



An

gle

(°)


Withers

Left Right

127

4.4 Discussion

The aim of this study was to determine if side reins have an effect on vertical

displacement amplitudes of the pelvis and withers of horses trotting on an aqua-

treadmill at increasing water depths. Prior to this study, it was observed that side

reins or some form of training aid are routinely used when working horses

overground so it was considered pertinent to also use some form of training aid

when working horses in an aqua-treadmill. It was also noted through experience

and observation of working with horses on an aqua-treadmill that often, the horse

would lose concentration in their work, raising their heads and altering their body

position instead of maintaining a relaxed and balanced head and neck position. It

was considered that the use of a training aid would be beneficial in helping to

maintain a more correct head and neck posture and side reins were deemed to be

the most straightforward, easy to use, and safe option within the aqua-treadmill

environment. All horses taking part in the study were regularly lunged in side

reins overground so the concept of side reins was not a new phenomenon to

them. However, it was the first time that any of the horses had been exercised on

an aqua-treadmill in side reins. The regular use of side reins in overground

training informed the decision to include side reins in the exercise protocol within

the aqua-treadmill in order to make some quantifiable recording and measurement

of the impact of side reins on movement patterns.

Ultimately it was determined in this study that the use of side reins had no

significant effect on the vertical displacement amplitudes achieved by either the

pelvis or the withers. No larger or no smaller vertical displacements were

recorded with the use of side reins. However, the same trends as seen previously

128

and reported in Chapter 3 were still observed – increasing water depth increased

vertical displacement amplitudes. This implies, as previously, that a greater depth

of water encourages the horse to work harder to raise themselves up over the

water therefore possibly building greater strength through muscular development.

It also implies that the use of side reins does not alter these vertical displacement

amplitudes and suggests that side reins have no effect in how hard the horse has

to work to achieve the greater amplitudes at greater depths. It could be suggested

however, that although no difference in vertical displacements were observed,

other muscles were perhaps engaged in order for comparative vertical

displacement amplitudes to be achieved both with and without the side reins. It

has previously been claimed in the literature that when working horses, lowering

the neck and working the horse downwards and forwards increases the movement

of the back and strengthens the back muscles (Denoix et al., 2001) and that the

hindlimbs become more engaged when the horse has a low head position where

the poll is at the same level or lower than the withers and with a wider head/neck

angle (Roepstorff et al., 2002). The results of this study perhaps concur with these

reports and in fact, by lowering the horses head and neck, greater engagement of

the back and hindlimb muscles are generated, along with the greater compression

of the limbs previously reported in Chapter 3 therefore working the horse harder to

create the vertical displacement required to work up and over the water. A further

study to investigate the effect of side reins on vertical displacement amplitudes

overground would be very beneficial.

As defined by the Federation Equestre Internationale (FEI) Rulebook for Dressage

2017, dressage desires lightness in the forehand and the engagement of the

hindquarters with a lively impulsion (Anon, 2017). Strasser (1913) cited in Slijper

129

(1946) and in Jeffcott (1979c) proposed the bow and string mechanism that is now

widely accepted which suggests that both the fore and hind limbs as well as the

abdominal and axial muscles are all involved in controlling trunk movement, and it

is well thought that a low head position and greater engagement of the hind

quarters is the best way to improve and develop equine trunk muscle (Cottriall et

al., 2009). In order for the hindlimbs to be engaged and the head and neck to

stretch downwards and forwards, the dorsal and ventral muscles must act to

stabilize the back against excessive flexion and extension as well as stabilising

lateral flexion and axial rotation (Cottriall et al., 2009). It has been suggested that

side reins, and other training aids, may aid in the activation and development of

these muscles to perform these stabilisations, however, when surface

electromyography (EMG) in the longissimus dorsi muscle was measured in horses

at trot on the lunge there was no significant effect of side reins on EMG activity but

there was a significantly greater EMG activity of the longissimus dorsi on the

inside of the circle (Cottriall et al., 2009). There is currently no information on

surface EMG activity of the longissimus dorsi when exercising horses in an aqua-

treadmill but as horses are maintained in a straight line in an aqua-treadmill it

would be very interesting to assess surface EMG in this way both with and without

side reins, and also, in both walk and trot as Cottriall et al. (2009) found surface

EMG the highest when walking on the lunge with no training aid (and again the

inside longissimus dorsi showed higher levels of activity than the outside) and

highest in the trot when only the hindquarter strap of a Pessoa training aid was

used (again inside muscle showed higher activity levels than the outside muscle).

Although not demonstrated in the EMG work by Cottriall et al. (2009) it is

recognised that side reins maintain a low position of the head and neck which

130

should ensure maximal hind limb engagement and development of the horse’s

topline muscles.

A later study specifically investigating the effect of the Pessoa Training Aid

(Walker et al., 2013) found that the Pessoa resulted in decreased stride length and

a decreased head angle and lowered head and neck position which agreed with

the findings of the Cottriall et al. (2009) study with both the Pessoa and side reins

where a shorter stride length and shorter neck length was noted. The shortened

neck and shortened stride length is most likely due to an influence of pressure on

the bit. In the present study, it is suggested that the element of water in the aqua-

treadmill goes somewhere to counteract the shorter stride lengths seen in previous

studies when training aids are applied as previous studies investigating the aqua-

treadmill have identified longer stride lengths and increased range of motions in

limb joints when exercising through water albeit these studies investigated walk,

not trot (Scott et al., 2010; Mendez-Angulo et al., 2013). A further consideration to

the counteraction of a potential shortened stride length with the addition of side

reins is that of the weight of the water and the tactile stimulation of the water

where studies have shown that the addition of weight to the pasterns along with

tactile stimulation of the pasterns resulted in a significant increase in the flexion of

the fetlock, tarsal and stifle joints when trotting (Clayton et al., 2010; Clayton et al.,

2011). Water could be considered as both a weight and a tactile stimulator

thereby increasing range of motion in the limb joints. Even small changes in digital

mass may have significant effects on kinematics due to the rapid accelerations

experienced during the gait cycle (Back et al., 1995a).

131

It is likely that the ventral muscles such as the abdominals and pectorals and

deeper muscles of the back become more engaged when exercising within an

aqua-treadmill when the horse must control their balance and rhythm on the

treadmill belt and work to lift themselves up over the water. Working over the

water likely requires a greater amount of muscle engagement as demonstrated by

the greater vertical amplitudes achieved in deeper water. Plus, the resistance of

the water creates a weight that also should be counteracted by a greater muscle

engagement. Further studies must be conducted to assess and determine if

exercising in side reins in the aqua-treadmill effects the overall workload of the

exercise session and where these changes in muscle engagements occur.

Indeed, a study that determines the workload of individual muscles when

exercising through different depths of water would also be highly relevant.

Previous studies have also identified that head and neck position have an effect

on stride length and the kinematics of the back (Faber et al., 2002; Johnston et al.,

2002; Rhodin et al., 2005; Gomez Alvarez et al., 2006; Rhodin et al., 2009).

Interestingly, in the Rhodin et al. (2005) study, stride length was shortest with the

head restricted in the highest position in the walk, but in trot, stride length was

independent of head and neck position, but it has been previously determined that

movement of the back is related to stride length where horses with longer strides

extend and flex their backs in the caudal saddle region to a greater extent at the

walk (Johnston et al., 2002) and an increasing stride length was correlated with an

increasing flexion/extension range of movement for most of the vertebrae at both

walk and trot (Faber et al., 2002; Rhodin et al., 2005). Elevating the head and

neck has been found to lead to extension in the cranial part of the spine and

flexion in the caudal part and lowering the head and neck has the opposite effect

132

in the unridden horse when trotting on a treadmill (Gomez Alvarez et al., 2006).

Unfortunately, no further parameters of the back were investigated in this present

study, but a consideration to the flexion/extension of the back when using side

reins in the aqua-treadmill would give scope for further investigation.

Anecdotally, riders and trainers are impressed with muscle developments in

horses that are regular uses of an aqua-treadmill which are displayed by an

apparent better core strength, stability and balance. This study only investigated a

single parameter of two bony anatomical landmarks moving in one plane to assess

the effect of water depth and side reins on vertical displacement amplitudes, but

there are of course, many integral essential body systems that contribute to

understand these findings. Side reins may change the shapes a horse makes with

regards to the positioning of the head and neck but have no apparent effect on the

workload in terms of vertical displacements achieved in the pelvis and withers

even at increasing water depths, so it may have been expected that there were

some changes in the results of the pitch data. There was no effect of side reins on

the pitch amplitudes throughout a stride cycle in both the pelvis and withers.

Side reins may not be the only training aid to act in this manner in the aqua-

treadmill, although as previously discussed, the training aid used needs to have

further safety considerations than would be considered overground due to the

nature of the confined environment within the aqua-treadmill – it is not possible to

get to the horse quickly to free them from a line trapped around a leg for example.

For this reason, it was considered that training aids such as a Pessoa are not

suitably safe to be used within an aqua-treadmill. However, there are a number of

other training aids that may be utilised on the aqua-treadmill that produce similar

133

or maybe more significant results than side reins owing to their different actions

that may be more suitable to different horses. Other potential possible training

aids include; an elastic bungie (Old Mill Saddlery, 2017), a Kavalkade HO

Lungeing Aid (Sydney Free Saddlery, 2017), a Chambon (Equestrian and Horse,

2017a), or a de Gogue (Equestrian and Horse, 2017b). Further consideration

must also be given to how side reins effect the vertical displacement amplitudes of

the pelvis and withers when trotting overground.

4.5 Conclusion

Ultimately, side reins appear to have no effect on vertical displacement amplitudes

of the pelvis or withers when trotting on an aqua-treadmill at increasing water

depths. It could be recommended or suggested from this study that it is

advantageous to work horses in side reins or another safely attachable training aid

(that does not directly interfere with the limb movement of the horse). It can be

concluded that the use of side reins do not appear to affect how hard a horse is

working including at different water depths as the vertical displacement amplitudes

achieved were similar to those achieved when working without side reins.

Therefore, the horse is perhaps not exerting any more energy or working any

harder, but it could be argued that the side reins are maintaining a more correct

head and neck position so that the horse may be working more constructively over

the back and aiding the development of the topline muscles along with engaging

the abdominal muscles for correct posture and balance plus engaging the

hindquarters. Repeated use of the aqua-treadmill with or without side reins would

perhaps be beneficial in promoting the engagement of the hindquarters, the

134

lightness of the forehand and the strengthening and development of the abdominal

and back muscles.

135

CHAPTER 5: Mediolateral displacements of the equine pelvis and withers when trotting on the aqua-treadmill

5.0 Introduction

The equine vertebral column is well documented in anatomical texts (Jeffcott

1979a; Jeffcott 1979b; Goody and Goody, 2000). Biomechanical analysis is

traditionally carried out during treadmill locomotion and often in relation to

investigation of back pain, lameness or poor performance (Butler et al., 2000;

Faber et al., 2000; Faber et al., 2001a, 2001b, 2001c; Johnston et al., 2002; Faber

et al., 2002; Erichson, 2003; Wennerstrand et al., 2004; Johnston et al., 2004;

Erichson et al., 2004; Barrett et al., 2006; Van Weeren et al., 2010; Allen et al.,

2010; Findley and Singer, 2015; Burns et al., 2016). Recent advances in

technology have enabled more detailed investigation into the movement of

anatomical structures during locomotion overground (Audigié et al., 1999) which

can provide a greater insight into abnormalities that may be altered or difficult to

detect with treadmill locomotion. Buchner et al. (1994a) has documented

differences in treadmill versus overground locomotion such as an increased

stance phase of the forelimbs and an increase in caudal movement during the

retraction phase of both fore and hind limbs on the treadmill compared to

overground. These studies are important in aiding understanding of

thoracolumbar biomechanical changes during exercise and may provide a

baseline for comparison when investigating biomechanical changes during aqua-

treadmill exercise.

136

Aqua-treadmill exercise has been shown to alter the movement pattern of the

limbs during walk (Scott et al., 2010; Mendez-Angulo et al., 2013; Lefrancois and

Nankervis, 2016). As limb and back movement are dynamically linked (van

Weeren, 2009; van Weeren et al., 2010) there is an expectation that changes in

water depth will impact back kinematics where limbs are influenced. Three major

movements of equine intervertebral joints have been described (Jeffcott, 1980;

Townsend et al., 1983; Clayton & Townsend, 1989; Denoix, 1999; van Weeren,

2009):

• flexion and extension movements occurring in the median plane around a

transverse axis,

• lateral flexion to the left and right sides in the horizontal plane around a

dorsoventral axis,

• left or right rotation occurring around a longitudinal axis.

It is only in the last five years that changes in the vertebral column have been

reported when walking through water (Mooij et al., 2013; Nankervis et al., 2016).

To date it seems that only walk has been investigated during aqua-treadmill

locomotion despite the treadmills being capable of safely imitating speeds of up to

18kmh which is more than adequate for most horses to achieve a ‘working trot’.

Ridden trot speeds have been documented as ranging from 11.5km/h in collected

trot, 13km/h in working trot, 16.1km/h in medium trot and 17.8km/h in extended trot

(Clayton, 1994a) while trotting in hand has been reported to be approximately

14.25km/h (Galisteo et al., 1998). Mooij et al. (2013) used videography to assess

pelvic rotation, axial rotation (roll) and lateral bend in horses walking at increasing

water depths and determined that pelvic flexion increased with water depth which

subsequently leads to increased stride length (Scott et al., 2010). Lateral bend

decreased with increasing water depth (Mooij et al., 2013) indicating that the

137

increased resistance of water leads to a compensatory pattern of movement.

Axial rotation increased when the water was at carpus level, but as the water

deepened to elbow and shoulder depths, axial rotation decreased. This is thought

to be as the horses were no longer able to step over the water forcing them to

adapt their movement as a result of the resistance of the water which lead to

increased pelvic flexion and reduced lateral bending (Mooij et al., 2013).

Nankervis et al. (2016) utilised a sophisticated three-dimensional gait analysis

system (Qualisys©) to investigate flexion and extension of the vertebral column

whilst horses were walking on an aqua-treadmill at increasing water depths. They

determined that deeper water increases the flexion-extension range of motion

when compared to walking at very low water levels; however, very deep water

(elbow height) resulted in increased mid thoracic extension. Whilst no effect of

water depth on pelvic displacement was found which contradicts the earlier study

where significant changes in axial rotation and pelvic displacement were seen

(Mooij et al., 2010). It is suggested in the most recently published study by

Nankervis et al. (2016), that the absence of a trend in pelvic displacement and

subsequent axial rotation with increasing water depths is indicative of the horses

adopting different hindlimb movement strategies. This change in movement is

possibly due to differences in muscle strength and activation levels required

suggesting possible contraindications if an inappropriate water depth is utilised for

a specific horse.

Using the main methodologies already detailed in Chapter 2, the aim of this

chapter was to quantify mediolateral displacement of the withers and pelvis to

identify an increase or decrease in lateral flexion with increasing water depths. It

138

is understood that this is not perhaps lateral flexion in the true sense as the horses

are moving in a straight line on the treadmill belt, but assessing the total deviation

from midline through a stride cycle and the subsequent amplitude of movement in

the horizontal plane representing range of motion. Further, this study aimed to

investigate the change in mediolateral angles and roll around the midline in order

to give an overall impression of changes in equine vertebral joints in the horizontal

plane and longitudinal axis when exercising through increasing water depths from

mid P3 to mid carpus.

5.1 Aims

The aims of this chapter were to investigate the effect of water depth and side

reins on mediolateral displacement of the pelvis and withers; lateral flexion of the

spine and roll of the pelvis and withers. Specifically:

1. Analyse the mediolateral displacement amplitudes of the pelvis and withers

when trotting on the aqua-treadmill at increasing water depths.

a. Determine the effect of side reins on these displacements, including

any effect on symmetry

2. Analyse the roll of the pelvis and withers when trotting on the aqua-treadmill

at different water depths.

a. investigate the effect of side reins on the roll

3. Analyse the change in mediolateral flexions from the withers to T13 to the

pelvis at increasing water depths.

139


5.2.1 Mediolateral displacement amplitudes

The methods and protocol used to collect the data were the same as previously

described in Chapter 2. A custom written script was generated for Matlab© to this

time extract the displacement data along the x-axis to calculate mediolateral

displacement amplitudes rather than vertical displacement amplitudes along the z-

axis. Figure 5.1 shows the orientation of the axes in the inertial measurement

units.

Figure 5.1: Orientation of the axes in a triaxial inertial measurement unit. Showing also the correct orientation of the unit on a horse (although the horse in the image has hemispherical light reflective markers attached for the optical motion capture methodology). The z-axis has been previously utilised in this project when researching vertical displacement amplitudes. And the x-axis is now being used in this chapter to investigate mediolateral displacement amplitudes. In all cases the y-axis denotes forward movement on the aqua-treadmill.

X

Y

Z

140

Mediolateral displacements were calculated in the same initial way as for vertical

displacements with regards to cutting the raw data into individual strides according

to left hind stance patterns with a double numerical integration from acceleration to

calculate displacements. Strides were cut using the minima of vertical

displacement at left hind stance. Only complete whole strides were utilised in the

analysis. Mediolateral displacements are presented as an approximate single

sinusoidal pattern for each stride with deviation away from the midline (either

positive or negative) calculated by establishing the distance deviated from a

central point which was considered as the sensor at T13 according to previous

literature stating that T13 most closely corresponds to the body centre of mass

movement (Buchner et al., 2000; Warner et al., 2010). The position of the sensor

at T13 was then subtracted from the position of the sensors either at the pelvis or

withers to determine movement to the left (negative) and movement to the right

(positive) thereby giving the relative positions of both the pelvis and withers. Total

mediolateral displacement amplitudes were then able to be calculated by

subtracting the relative position of T13 for the withers and pelvis.

5.2.2 Roll amplitudes

To investigate the roll of each of the inertial measurement units on the pelvis and

withers, a custom written Matlab© script was used to extract the data from the

gyroscopes.

141

5.2.3 Mediolateral flexions

Using the sophisticated QTM Track Manager© software it was possible to

investigate the angle data between the labelled points of the withers, mid back and

pelvis. There were six horses that had a mid-back (T13) light reflective marker

attached for investigation. The change in angle between these points was

calculated at the increasing water depths and a repeated measures one-way

ANOVA was used to test for significance. Figure 5.2 illustrates the mediolateral

flexions of the spine with T13 as the central point.

142

Figure 5.2: An illustration of how the anatomical landmarks of the withers (T4/5), mid back (T13) and pelvis (tuber sacrale) form a straight line and at left hind stance (LHS) there is flexion to the left which changes to a flexion to the right as the horse changes to a right hind stance (RHS).

Straight 180°

RHS >180°

LHS <180°

143

5.2.4 Statistical Analyses








Statistics, 2015).

5.3 Results

There were a total of ten complete sets of data for the analyses on mediolateral

displacements and roll amplitudes (Chapters 5.3.1 and 5.3.2). Horses heights

ranged from 147-180cm (mean ± SD 163.20 ± 9.84) and ages ranged from 8-19

years (mean ± SD 12.88 ± 3.84). For the mediolateral flexions analysis (Chapter

5.3.3) there were six sets of data with heights ranging from 157-172cm (mean ±

SD 163.8 ± 4.956) and ages ranging from 11-13 years (mean ± SD 12.00 ± 0.894).

144

5.3.1 Mediolateral Displacement Amplitudes


pelvis or withers, side reins and water depth on the mean mediolateral

displacement amplitudes of ten horses trotting on the aqua-treadmill at increasing

water depths; the means were plotted (Figure 5.3).

There was no significant three-way interaction between ‘pelvis or withers’, ‘water

depth’ or ‘side reins’ (F(3,27) = 0.208, p = 0.326). There was a statistically

significant two-way interaction between ‘pelvis or withers’ and ‘water depth’ (F(3,27)

= 8.819, p < 0.001) which required further analysis.

Post hoc analyses of the simple main effect of pelvis versus withers (Table 5.1)

determined that the withers had significantly greater mediolateral displacements

than the pelvis without side reins at a water depth of mid P3 (F(1,9) = 9.036, p =

0.015), a mean (±SEM) increase in the withers of 15.49 (±5.15) mm; and with side

reins at a water depth of mid P3 (F(1,9) = 29.917, p < 0.001), a mean (±SEM)

increase in the withers of 21.38 (±3.91) mm. At a water depth of mid fetlock, the

withers had greater mediolateral displacement amplitudes than the pelvis without

side reins (F(1,9) = 7.177, p = 0.025), a mean (±SEM) increase in the withers of

16.81 (±6.28) mm.

Post hoc analyses of the simple main effect of water depth (Table 5.2) determined

that in the pelvis, there was a significant increase in mediolateral displacement

amplitudes with the use of side reins between the depths of mid P3 and mid

145

carpus (F(3,27) = 5.192, p = 0.006) with a mean (±SEM) increase of 12.73 (±2.64)

mm.

Post hoc analyses of the simple main effect of water depth determined that in the

withers, (Table 5.2), there was a trend of decreasing mediolateral displacement

amplitudes with increasing water depth both without (F(3,27) = 7.806, p = 0.001) and

with side reins (F(3,27) = 6.673, p = 0.009), with the significant differences laying

between the depths of mid P3 and mid carpus without side reins (p = 0.046), with

a mean (±SEM) decrease of 10.51 (±3.07) mm, and between mid fetlock and mid

carpus without side reins (p = 0.003), with a mean (±SEM) decrease of 9.89

(±1.90) mm. With the use of side reins in the withers, the significant differences

lay between mid P3 and mid carpus (p = 0.011), a mean (±SEM) decrease of 8.66

(±1.99) mm, and between mid fetlock and mid carpus (p = 0.012), with a mean

(±SEM) decrease of 8.60 (±2.01) mm.

146

Table 5.1: Post hoc analysis of the significant simple main effect of pelvis versus withers on mediolateral displacement amplitudes of 10 horses trotting on the aqua-treadmill at increasing water depths. Pairwise comparison with Bonferroni correction shown. Mean difference shown in mm.


Mean difference

Standard error

Sig.a Lower Bound

Upper Bound

NO

SID

E R

EIN

S

Mid P3 Pelvis Withers -15.49* 5.15 0.015 -27.15 -3.83

Mid Fetlock

Pelvis Withers -16.81* 6.28 0.025 -30.99 -2.62

Mid MC3 Pelvis Withers -10.99 5.93 0.097 -24.41 2.43

Mid Carpus

Pelvis Withers -0.94 6.37 0.885 -15.36 13.47

YES

SID

E R

EIN

S

Mid P3 Pelvis Withers -21.38* 3.91 <0.001 -30.22 -12.54

Mid Fetlock

Pelvis Withers -14.95 7.50 0.077 -31.92 2.02

Mid MC3 Pelvis Withers -6.16 8.17 0.470 -24.64 12.31

Mid Carpus

Pelvis Withers 0.02 5.19 0.997 -11.72 11.75


147

Table 5.2: Post hoc analysis of the significant simple main effect of water depth on mediolateral displacement amplitudes of the pelvis and withers of 10 horses trotting on the aqua-treadmill at increasing water depths. Pairwise comparison with Bonferroni correction shown. Mean difference shown in mm.


Mean difference

Standard error

Sig.a Lower Bound

Upper Bound

PEL

VIS

NO

SID

E R

EIN

S

Mid P3 Mid Fetlock 1.94 5.89 1.000 -17.88 21.75

Mid MC3 0.68 3.87 1.000 -12.36 13.70

Mid Carpus -4.04 3.77 1.000 -16.73 8.46

Mid Fetlock Mid P3 -1.94 5.89 1.000 -21.75 17.88

Mid MC3 -1.26 4.33 1.000 -15.82 13.30

Mid Carpus -5.98 5.21 1.000 -23.50 11.55

Mid MC3 Mid P3 -0.68 3.87 1.000 -13.71 12.36

Mid Fetlock 1.26 4.33 1.000 -13.30 15.82

Mid Carpus -4.72 1.54 0.082 -9.91 0.48

Mid Carpus Mid P3 4.04 3.77 1.000 -8.64 16.73

Mid Fetlock 5.98 5.21 1.000 -11.55 23.50

Mid MC3 4.72 1.54 0.082 -0.48 9.91

YES

SID

E R

EIN

S

Mid P3 Mid Fetlock -6.36 4.07 0.912 -20.04 7.31

Mid MC3 -11.44 3.83 0.091 -24.31 1.43

Mid Carpus -12.73* 2.64 0.006 -21.63 -3.84

Mid Fetlock Mid P3 6.36 4.07 0.912 -7.31 20.04

Mid MC3 -5.08 4.55 1.000 -20.37 10.21

Mid Carpus -6.37 2.27 0.122 -13.99 1.25

Mid MC3 Mid P3 11.44 3.83 0.091 -1.43 24.31

Mid Fetlock 5.08 4.55 0.000 -10.21 20.37

Mid Carpus -1.29 3.66 0.000 -13.60 11.03

Mid Carpus Mid P3 12.73* 2.64 0.006 3.84 21.63

Mid Fetlock 6.37 2.27 0.122 -1.25 13.99

Mid MC3 1.29 3.66 1.000 -11.03 13.60

WIT

HER

S

NO

SID

E R

EIN

S

Mid P3 Mid Fetlock 0.62 2.08 1.000 -6.36 7.60

Mid MC3 5.18 2.25 0.283 -3.40 12.76

Mid Carpus 10.51* 3.07 0.046 0.17 20.84

Mid Fetlock Mid P3 -0.62 2.08 1.000 -7.60 6.36

Mid MC3 4.56 2.32 0.484 -3.23 12.35

Mid Carpus 9.89* 1.90 0.003 3.48 16.29

Mid MC3 Mid P3 -5.18 2.25 0.283 -12.76 2.40

Mid Fetlock -4.56 2.32 0.484 -12.35 3.23

Mid Carpus 5.33 2.94 0.619 -4.56 15.21

Mid Carpus Mid P3 -10.51* 3.07 0.046 -20.84 -0.17

Mid Fetlock -9.89* 1.90 0.003 -16.29 -3.48

Mid MC3 -5.33 2.93 0.619 -15.21 4.56

YES

SID

E R

EIN

S

Mid P3 Mid Fetlock 0.06 1.87 1.000 -6.22 6.34

Mid MC3 3.77 3.08 1.000 -6.61 14.15

Mid Carpus 8.66* 1.99 0.011 1.98 15.35

Mid Fetlock Mid P3 -0.06 1.87 1.000 -6.34 6.22

Mid MC3 3.71 1.53 0.231 -1.45 8.86

Mid Carpus 8.60* 2.01 0.012 1.83 15.37

Mid MC3 Mid P3 -3.77 3.08 1.000 -14.15 6.61

Mid Fetlock -3.71 1.53 0.231 -8.86 1.45

Mid Carpus 4.89 2.60 0.556 -3.86 13.65

Mid Carpus Mid P3 -8.66* 1.99 0.011 -15.35 -1.98

Mid Fetlock -8.60* 2.01 0.012 -15.37 -1.83

Mid MC3 -4.89 2.60 0.556 -13.65 3.86


148

Figure 5.3: Mean (±SEM) mediolateral displacement amplitudes (in millimetres) of the equine pelvis (left) and withers (right) when trotting on an aqua-treadmill at increasing water depths both without side reins (orange) and with side reins (green) (n=10). The withers had significantly larger mediolateral displacements than the pelvis at mid P3 without side reins * (p < 0.05), and with side reins # (p < 0.01). The withers had significantly larger mediolateral

displacements than the pelvis at mid fetlock without side reins (p < 0.05). In the pelvis with side reins there was a significant difference in mediolateral displacements between mid P3 and mid carpus a (p < 0.01). In the withers there was a trend of decreasing mediolateral displacements with increasing water depth, with a significant difference between mid P3 and mid carpus without side reins b (p < 0.05), and between mid fetlock and mid carpus without side reins c (p < 0.01); with side reins there was a significant difference between mid P3 and mid carpus d (p < 0.05), and between mid fetlock and mid carpus e (p < 0.05).

*

a#

a

0

10

20

30

40

50

60

70



Dis

pla

cem

en

t (m

m)


Pelvis

b*

d#

c

e

bc

de

0

10

20

30

40

50

60

70



Dis

pla

cem

en

t (m

m)


Withers

149

5.3.2 Roll Amplitudes

The second element of this chapter investigated the roll of the inertial

measurement units of the pelvis and withers when trotting on the aqua-treadmill at

increasing water depths and the means were plotted (Figure 5.4).


pelvis or withers, side reins and water depth on the mean roll amplitudes of ten

horses trotting on the aqua-treadmill at increasing water depths. There was no

statistically significant three-way interaction between ‘pelvis or withers’, ‘water

depth’ or ‘side reins’ (F(3,27) = 0.695, p = 0.458). There was a statistically

significant two-way interaction between ‘pelvis or withers’ and ‘side reins’ (F(1,9) =

8.535, p = 0.017) which required further analysis. Post hoc analyses determined

that when the water depth was at mid P3, the withers had significantly greater

mean (±SEM) roll amplitudes than the pelvis both without side reins at 6.47

(±1.64)° (F(1,9) = 15.618, p = 0.003), and with side reins at 7.34 (±3.01)° (F(1,9) =

5.949, p = 0.037) (Table 5.3). Also, when the water depth was at mid fetlock, the

withers had significantly greater roll amplitudes than the pelvis both without side

reins at 9.21 (±3.38)° (F(1,9) = 7.410, p = 0.024), and with side reins at 6.30

(±2.54)° (F(1,9) = 6.162, p = 0.035) (Table 5.3).

Post hoc analysis on the simple main effect of side reins determined that there

was a significant decrease in mean (±SEM) roll amplitude of 3.34 (±1.34)° in the

withers at a water depth of mid fetlock when the side reins were attached (F(1,9) =

6.230, p = 0.034) (Table 5.4). There was no effect of water depth on roll

amplitudes.

150

Table 5.3: Post hoc analysis of the significant simple main effect of pelvis versus withers on roll amplitudes of the pelvis and withers when trotting on an aqua-treadmill at increasing water depths. Pairwise comparison with Bonferroni correction shown. Mean difference shown in degrees.



difference Standard

error Sig.a

Lower Bound

Upper Bound

NO Side Reins

Mid P3 -6.47* 1.64 0.003 -10.18 -2.77

Mid Fetlock -9.21* 3.38 0.024 -16.86 -1.56

Mid MC3 -6.69 3.18 0.065 -13.88 0.50

Mid Carpus -5.34 3.26 0.136 -12.71 2.03

YES Side Reins

Mid P3 -7.34* 3.01 0.037 -14.14 -0.53

Mid Fetlock -6.30* 2.54 0.035 -12.03 -0.56

Mid MC3 -6.01 2.82 0.061 -12.39 0.36

Mid Carpus -4.36 2.82 0.156 -10.73 2.01


Table 5.4: Post hoc analysis of the significant simple main effect of side reins on roll amplitudes of the pelvis and withers when trotting on an aqua-treadmill at increasing water depths. Pairwise comparison with Bonferroni correction shown. Mean difference shown in degrees.


No Side Reins to Yes Side Reins Mean

difference Standard

error Sig.a

Lower Bound

Upper Bound

PELVIS

Mid P3 -0.42 0.42 0.344 -1.38 0.54

Mid Fetlock 0.43 0.62 0.509 -0.97 1.82

Mid MC3 -0.25 0.58 0.675 -1.56 1.06

Mid Carpus -0.47 0.88 0.605 -2.48 1.53

WITHERS

Mid P3 -1.29 2.83 0.660 -7.68 5.11

Mid Fetlock 3.34* 1.34 0.034 0.31 6.36

Mid MC3 0.42 1.23 0.738 -2.35 3.20

Mid Carpus 0.51 1.45 0.735 -2.78 3.79


151

Figure 5.4: Mean (±SEM) roll amplitudes (in degrees) throughout a stride cycle of the inertial measurement units stationed at the pelvis (left) and withers (right) when trotting on an aqua-treadmill at increasing water depths both without side reins (orange) and with side reins (green) (n=10). The withers had significantly greater roll amplitudes than the pelvis at a depth of mid P3 without side reins a (p < 0.01) and with side reins b (p < 0.05). The withers had significantly greater roll amplitudes than the pelvis at a depth of mid fetlock both without and with side reins c (p < 0.05). In the withers at the depth of mid fetlock there was a significant decrease in roll amplitudes when side reins were applied d (p < 0.05). There was no effect of water depth on roll amplitudes.

a b cc

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30



An

gle

(d

egr

ee

s)


Pelvis

a

b

cd

cd

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30



An

gle

(d

egr

ee

s)


Withers

152

5.3.3 Mediolateral Flexions

The third element of this chapter, involved the investigation of the mediolateral

flexions from the withers through T13 to the pelvis. Mean flexions were plotted

(Figure 5.5).

A one-way repeated measures ANOVA was conducted to determine the effects of

water depth on the change in mediolateral angle through withers-T13-pelvis of six

horses trotting on the aqua-treadmill at increasing water depths. Changes in the

water depth did not elicit statistically significant changes in the change in

mediolateral spine angle (F(3,15) = 0.426, p = 0.737, partial η2 = 0.079).

153

Figure 5.5: Mean (±SEM) change in mediolateral angle of the equine spine (Angle XY, Withers - T13 - tuber sacrale) when trotting on the aqua-treadmill at increasing water depths (n=6). There was no significant effect of water depth on mediolateral spine flexions.

0

1

2

3

4

5

6

7

8


An

gle

(d

egre

es)

Water Depth

Equine Spine Angle(withers - T13 - tuber sacrale)

154

5.4 Discussion

The aims of this study were to determine if water depth has a significant effect on

how the axial skeleton moves mediolaterally in a horizontal plane around a

dorsoventral axis, including investigating any effect of side reins, investigating the

roll of the inertial measurement units and investigating the change in mediolateral

angle throughout a stride cycle. No studies have appeared to investigate these

parameters previously with a horse exercising on an aqua-treadmill at trot.

5.4.1 Mediolateral displacement amplitudes

Mediolateral displacement amplitudes throughout the stride were found not to

change with increasing water depths. This is similar to a previous study where

lateral bending range of motion was found not to change significantly when the

horses were walking at a water depth of the fetlock or carpus but decreased

significantly when the water depth was increased to the levels of the elbow and

shoulder joints (Mooij et al., 2013). The present study did not take the water

above the level of the carpus and only included trot not walk, so no direct

comparison can be made; however, it seems likely that the mechanisms for

trotting and walking over a low level of water may be similar whereby the horse is

required to step up over the water. The Mooij et al. (2013) study suggests that at

the deepest water levels of elbow and shoulder depth, that the water provides

some stability as lateral bending was significantly decreased but pelvic flexion was

significantly increased so suggesting a change in the engagement of different

155

muscles to produce a different walking pattern. Pelvic flexion was not measured in

the current study so again, no direct comparison can be made.

It appeared from Figure 5.1 that overall, the withers had a greater overall

mediolateral displacement than the pelvis but this was only found to be significant

at the lower water depths of mid P3 without side reins (p = 0.015) and with side

reins (p < 0.001), and mid fetlock (p = 0.025). It is interesting that the withers

exhibited a greater mediolateral displacement than the pelvis at a lower water

depth suggesting that the front end of the horse is perhaps rocking from side to

side to jump over the water when the water is first introduced. This perhaps is

comparable to the evidence that the withers overall appeared to have lower overall

vertical displacement amplitudes than the pelvis suggesting that the front end of

the horse compensated the water by bending at the carpus, and so here

suggesting that the bending action of the carpus one leg at a time creates a

rocking action that actually sways the horse from side to side thereby creating

these greater mediolateral displacement amplitudes. At the deeper water depths

of mid MC3 and mid carpus, the mediolateral displacement amplitudes of the

pelvis and withers were almost the same. This perhaps suggests that the front

end of the horse is no longer rocking by simply bending at the carpus, but actually

having to create a jump up over the deeper water. This would also correspond

with the evidence from Chapter 3 where vertical displacement amplitudes in the

withers were found to increase with deeper water, so the horse is now jumping up

over the water creating greater vertical lift and less mediolateral sway. Overall, the

withers exhibited a trend where increasing water depth reduced mediolateral

displacement amplitudes. This again would suggest that there is an increased

156

compression of the joints in the forehand to create this lift, which again correlates

with the findings of Chapter 3.

There was no apparent overall effect of side reins on mediolateral displacement

amplitudes. It is suggested that side reins may have had a stabilizing effect on the

side to side sway seen in the withers at lower water depths, and although a

reduction in mediolateral displacement amplitudes was seen with the addition of

side reins at the lower water depths (Figure 5.1), this trend was not found to be

significant. In the pelvis, there was no effect of side reins at all on mediolateral

displacement amplitudes. The application of side reins did not have the effect of

reducing mediolateral displacement amplitudes and therefore perhaps creating

any kind of stabilizing effect. Likewise, the application of side reins did not

increase mediolateral displacement amplitudes which would have been

unexpected.

5.4.2 Roll amplitudes

There was one clear trend in the roll data, the withers overall showed greater

changes in roll than the pelvis (p = 0.039). As discussed in the Methods chapter

(Chapter 2), a specially constructed withers mount was used to affix the inertial

measurement unit to the withers due to previous studies stating the difficulty of

affixing inertial measurement units to this anatomical area (Pfau et al., 2005) due

to the movement of the skin over bony protuberances which is well documented

(Van Weeren and Barneveld, 1986; Van den Bogert et al., 1990; Van Weeren et

al., 1990a; Van Weeren et al., 1990b; Van Weeren et al., 1992a, 1992b). It is

157

therefore suspected that these greater roll amplitudes may well be due to the

movement of the skin over the withers.

There was no effect of water depth on the roll amplitudes. Roll amplitudes in the

pelvis remained fairly constant (between 16 – 18 degrees) with the increasing

water depths evidencing that neither water depth nor side reins had an impact on

these displacements. In the withers, roll amplitudes appeared more variable but

again, there was no effect of water depth or side reins. It could have been

suggested perhaps, that the addition of side reins may have decreased the

variability in the roll due to perhaps having an effect on stabilising the front end of

the horse but this was not seen. Roll amplitudes in the withers decreased slightly

as the water depth increased which tallies with the decrease also seen in

mediolateral displacements but this effect was not found to be significant. It is

suggested that the lack of change in roll seen in the pelvis may be due to the

location of placement of the inertial measurement unit on the tuber sacrale. It is

perhaps not expected that a large variation would be seen here due to the

anatomical location being fairly level, so perhaps not subject to large changes in

roll. However, a previous study investigating axial rotation in the walk at

increasing water depths found that axial rotation increased significantly at each

successive water depth from the hoof control level to a level of the shoulder joint

(Mooij et al., 2013). It is possible that the significant increases seen in the walk

and not the trot are due to the fundamental difference in the kinematics of the gaits

and not due to any effect of water (Haussler et al., 2001; Johnston et al., 2002;

Johnston et al., 2004). However, it must also be noted that it is documented that

horses do modify their gait mechanics to compensate for injury and pain (Cadiot

and Almy, 1924; Buchner et al., 1995; Buchner et al., 1996a, 1996b; Uhlir et al.,

158

1997; Denoix and Audigie, 2001; Weishaupt et al., 2004; Weishaupt et al., 2006;

Gomez Álvarez et al., 2007; Gomez Álvarez et al., 2008). Although all horses

were deemed non-lame by a Vet prior to taking part in the study, minute changes

in gait that may not have been detectable by eye overground may have been

exacerbated or accentuated in the aqua-treadmill and this could potentially have

resulted in changes in range of motion in the thoracolumbar spine without

producing detectable changes in the kinematics of the limbs, for example a slightly

increased range of motion of the thoracolumbar back, a slightly decreased range

of motion of the lumbosacral segment and rotational motion changes of the pelvis

which have been noted when inducing a slight hindlimb lameness overground in

walk and trot (Gomez Álvarez et al., 2008). This potentially may have had

individual effects on each horse in the present study thereby reducing the

likelihood of finding a trend in this roll data.

5.4.3 Mediolateral flexions

Mediolateral flexions were investigated along with the mediolateral displacement

amplitudes and roll data. There was no effect of water depth on the change in

mediolateral spinal angle. It is suggested that perhaps as water depth increased

there may have also been an increase in the change in that mediolateral angle but

there were no changes seen. This actually suggests that even with greater depths

of water the spine actually remains very stable and is not perhaps ‘over’ or ‘hyper’

flexed by the horse trotting in greater depths of water. This suggests that you can

continue to work the horse hard benefitting from the benefits of the deeper water

but without any detrimental effects to over lateral flexing of the spine. It does not

159

appear that these mediolateral angles have been investigated in any previous

study.

5.5 Conclusion

This chapter has identified the relationships between the mediolateral movements

and roll of the equine spine when trotting on an aqua-treadmill at increasing

depths of water. Water depth had no effect on mediolateral displacements either

in the pelvis or withers of horses trotting on an aqua-treadmill. At lower water

depths, the withers exhibited greater mediolateral displacements than the pelvis

suggesting a side to side sway motion was initiated. At deeper water depths, the

mediolateral displacements in the withers had reduced to a level comparable with

those of the pelvis suggesting that deeper water controls front end mediolateral

movement. Side reins had no effect on mediolateral displacement amplitudes or

on roll amplitudes of horses trotting on an aqua-treadmill at increasing water

depths.

Roll amplitudes were significantly greater in the withers than the pelvis. This is

likely due to the movement of the skin over the bony protuberance of the withers.

The lack of significance in the roll data suggests that roll is therefore not a

parameter that has any bearing on studies investigating biomechanical parameters

of the axial skeleton. Mediolateral flexions of the spine were not affected by water

depth or side reins. This suggests that the horse can be worked harder at greater

water depths without over stressing the mediolateral capabilities of the spine.

160

CHAPTER 6: A comparison of overground and aqua-treadmill locomotion

6.0 Introduction

It is understood that horses were first put on a treadmill for research into

locomotion by Persson in 1967 and the treadmill has facilitated major advances in

biomechanical research by allowing the integration of biomechanical, physiological

and biochemical data to be collected under controlled conditions (Clayton, 1989).

A state-of-the-art treadmill for use with cinematography was described by

Fredricson et al. in 1983 where he reported that reproducibility between strides

was good but that there was a reduction in stride length and stride frequency.

Since then, treadmills have been commonly used for research purposes due to the

ability to move the horse in a straight line for a repeated number of reproducible

successive strides. Also, biomechanical differences between treadmills and

overground locomotion have been well reported and discussed earlier in this

thesis (Fredricson et al., 1983; Leach and Drevemo, 1991; Barrey et al., 1993;

Buchner et al., 1994a, 1994b).

The concept of aqua-treadmills was first introduced into scientific literature in

1989, (Auer, 1989) but the effect of water on any biomechanical parameters was

not studied until 2010 when stride length and frequency were investigated with

changes in water depth (Scott et al., 2010). As yet, there does not appear to be

any reports making a direct comparison between overground locomotion and

locomotion through water on an aqua-treadmill. This study sought to investigate

161

the effects of a very low water depth of just a few centimetres on the locomotion

parameters previously investigated in this project.

6.1 Aims

This chapter aimed to make a comparison of the locomotion of the pelvis and

withers overground and on the aqua-treadmill with just a small amount of water, to

determine the effect a water depth of mid P3 has on the locomotory parameters

previously measured in this project. Specifically, to:

1. Analyse the vertical displacement amplitudes achieved by the pelvis and

withers when trotting overground

a. investigate the effect of side reins on these displacements, including

any effects on symmetry

2. Compare and analyse the vertical displacement amplitudes of the

overground work to the aqua-treadmill work (water at mid P3) and

investigate the effect of side reins

3. Analyse the pitch of the inertial measurement units at the pelvis and withers

when trotting overground.

4. Compare and analyse the pitch of the inertial measurement units at the

pelvis and withers when trotting overground to trotting on the aqua-treadmill

(water at mid P3).

162

5. Analyse the mediolateral displacement amplitudes at the pelvis, withers and

poll when trotting overground.

a. investigate the effect of side reins on these displacements

6. Compare and analyse the mediolateral displacement amplitudes of the

overground work to the aqua-treadmill work (water at mid P3) in the pelvis,

withers and poll.

a. investigate the effect of side reins

7. Analyse the roll of the pelvis, withers and poll when trotting overground.

a. investigate the effect of side reins on the roll

8. Compare and analyse the roll from the overground work to the aqua-

treadmill work (water at mid P3)

a. investigate the effect of side reins on these comparisons

163


For this investigation, only the data from the Xsens© data collection was utilised

as this included the overground work and the side reins work. Data were collected

and handled in the manner as described previously in Chapter 2 for the Xsens©

data.









Statistics, 2015).

164

6.3 Results

The effects of water on vertical and mediolateral displacement amplitudes of the

pelvis and withers when trotting on an aqua-treadmill have previously been

reported in Chapters 3, 4 and 5, but now a comparison can be made with values

obtained from trotting overground. The same ten horses that were exercised on

the aqua-treadmill using the Xsens© technology were also trotted overground on

tarmac in a straight line both without and with side reins. There was a total of 10

sets of Xsens© data to complete this analysis. Horses’ heights ranged from 147-

180cm (mean ± SD 163.20 ± 9.84) and ages ranged from 8-19 years (mean ± SD

12.88 ± 3.84).

6.3.1 Overground Vertical Displacements

Vertical displacements of the pelvis and withers in horses trotting overground were

measured and the means plotted (Figure 6.1). A repeated measures ANOVA

(2x2x2) was conducted to determine the effects of pelvis or withers, side reins and

left or right diagonal pair stride on the mean vertical displacement amplitudes of

ten horses trotting overground.

There was no significant three-way interaction between ‘pelvis or withers’, ‘side

reins’ or ‘left or right diagonal pair’ (F(1,9) = 0.023, p = 0.956). There was a

significant two-way interaction between ‘pelvis or withers’ and ‘side reins’ (F(1,9) =

7.844, p = 0.021) and a significant two-way interaction between ‘pelvis or withers’

and ‘left or right’ (F(1,9) = 6.837, p = 0.028) which required further analysis.

165

Post hoc analyses of the simple main effects of pelvis versus withers (Table 6.1),

showed that with the addition of side reins on the left diagonal pair, vertical

displacement amplitudes were significantly lower in the withers than the pelvis

(F(1,9) = 9.719, p = 0.012), a mean (±SEM) difference of 22.90 (±7.35) mm.

Post hoc analyses of the simple main effects of Side Reins (Table 6.2) showed

that in the withers, there was a significant decrease in vertical displacement

amplitudes with the addition of side reins in both the left (F(1,9) = 10.555, p = 0.010)

and right (F(1,9) = 11.281, p = 0.008) diagonal pairs; a mean (±SEM) difference on

the left of 23.18 (±7.13) mm, and on the right of 22.99 (±6.85) mm. Principally, the

mean displacement in the withers was less when side reins were added. Side

reins had the effect of reducing displacement in the withers but not in the pelvis.

166

Table 6.1: Post hoc analysis of the significant simple main effect of pelvis versus withers on vertical displacements in horses trotting overground (n=10). Pairwise comparison with Bonferroni correction shown. Mean difference shown in millimetres.



difference Standard

error Sig.a

Lower Bound

Upper Bound

NO Side Reins Left 1.81 2.52 0.491 -3.90 7.52

Right -3.90 3.57 0.304 -11.98 4.19

YES Side Reins Left 22.90* 7.35 0.012 6.28 38.52

Right 17.06 9.47 0.105 -4.36 38.47

Based on estimated marginal means a Bonferroni adjustment for multiple comparisons * The mean difference is significant at the 0.05 level Table 6.2: Post hoc analysis of the significant simple main effect of side reins on vertical displacements of the pelvis and withers in horses trotting overground (n=10). Pairwise comparison with Bonferroni correction shown. Mean difference shown in millimetres.


No Side Reins to Yes Side Reins Mean

difference Standard

error Sig.a

Lower Bound

Upper Bound

PELVIS Left 2.09 1.17 0.107 -0.55 4.73

Right 2.04 1.95 0.323 -2.37 6.44

WITHERS Left 23.18* 7.13 0.010 7.04 39.32

Right 22.99* 6.85 0.008 7.51 38.47


167

Figure 6.1: Mean (±SEM) vertical displacement amplitudes (in mm) of the equine pelvis (left) and withers (right) when trotting in a straight line overground as informed by the Xsens© data (n=10), both with and without side reins on left (green) and right diagonal (pink) stride pair. No side reins indicated by dots. Side reins indicated by dashes. Vertical displacement amplitudes were significantly lower in the withers than the pelvis on the left diagonal pair with side reins a (p < 0.05). In the withers vertical displacement amplitudes were significantly less when side reins were added on both the left b (p < 0.05), and right diagonal pair c (p < 0.01).

a

0

10

20

30

40

50

60

70

80

NO side reins YES side reins

Overground

Dis

pla

cem

en

t (m

m)

Pelvis

Left Right

b

ab

c

c

0

10

20

30

40

50

60

70

80

NO side reins YES side reins

OvergroundD

isp

lace

me

nt

(mm

)

Withers

Left Right

168

6.3.1.a Overground Vertical Displacements – Symmetry

Considering the apparent effect of side reins creating asymmetry in the withers in

the previous analysis (6.3.1), it was worth studying the effects of the side reins on

the symmetry of the vertical displacements when the horse is trotting overground

in more detail. Figure 6.2 is created from the raw data and outlines the difference

in displacement of the left and right diagonal pair in the pelvis and withers both

with and without side reins, appearing to show that symmetry is increased in the

withers when side reins are applied.

A two-way repeated measures ANOVA showed no statistically significant two-way

interaction between pelvis or withers and side reins (F(1,9) = 2.726, p = 0.133).

There was no significant difference in the main effect of pelvis or withers (F(1,9) =

0.545, p = 0.479) and there was no significant difference in the main effect of side

reins (F(1,9) = 1.884, p = 0.203). Adding side reins made no significant difference

to the symmetry of the horses trotting overground.

169

Figure 6.2: Mean (±SEM) difference (mm) in vertical displacement amplitudes between left and right diagonal pair in the pelvis and withers without side reins (orange) and with side reins (green) (n=10). There was no effect of side reins on the vertical displacement symmetries of horses trotting overground.

0

1

2

3

4

5

6

7

8

9

10

NO SR YES SR NO SR YES SR

PELVIS WITHERS

Dif

fere

nce

in D

isp

lace

me

nt

(mm

)

Symmetry

170

6.3.2 Comparison between overground and ATMP3 vertical displacements

A comparison could be made between the vertical displacements of the pelvis and

withers trotting overground to the vertical displacements when the same ten

horses trotted on the aqua-treadmill at a low water depth (depth of mid P3)

(reported in Chapters 3 and 4) to determine what effect, if any, a very low depth of

water has on vertical displacement amplitudes. The means were plotted in Figure

6.3.

A repeated measures ANOVA (2x2x2x2) was conducted. Results of a four-way

analysis of variance are vast and take some deciphering but ultimately, the

significant elements of the analysis are reported as follows.

There was no statistically significant four-way interaction between the variables

(F(1,9) = 0.067, p = 0.801). But there was one statistically significant three-way

interaction between pelvis or withers, overground or ATMP3, and side reins (F(1,9)

= 5.923, p = 0.038).

Post hoc analysis of simple two-way interactions, simple simple main effects and

pairwise comparisons (Table 6.3) showed that trotting on the aqua-treadmill at a

water depth of mid P3 showed significantly greater vertical displacements in the

pelvis both without and with side reins than trotting overground both without and

with side reins (p < 0.001). In the pelvis, mean (±SEM) vertical displacements

were 28.18 (±4.51) mm higher on the aqua-treadmill than overground, without side

reins, and 32.01 (±3.66) mm higher with side reins. In the withers, trotting on the

aqua-treadmill at a water depth of mid P3 with side reins had significantly greater

171

vertical displacements than trotting overground with side reins (p = 0.003) a mean

difference of 29.46 (±7.46) mm, but there was no significant effect without side

reins.

Trotting on the aqua-treadmill at a water depth of mid P3 both without and with

side reins showed significantly lower vertical displacements in the withers than the

pelvis (p < 0.001) (as reported in Chapter 3 – the pelvis displaces higher than the

withers). When trotting overground with side reins, the pelvis exhibited

significantly greater vertical displacements than the withers (p = 0.012) and the

application of side reins had the significant effect of reducing vertical

displacements in the withers (p = 0.010) (also reported in chapter 6.3.1).

Table 6.3: Post hoc analysis of the significant simple two-way interactions and simple simple main effects for the comparison in vertical displacements of the pelvis and withers of horses trotting overground and on an aqua-treadmill at water depth of mid P3. Pairwise comparisons with Bonferroni correction shown. Mean difference shown in millimetres.



difference Standard

error Sig.a

Lower Bound

Upper Bound

OG NO SR 1.81 2.52 0.491 -3.90 7.52

YES SR 22.90* 7.35 0.012 6.28 39.52

ATMP3 NO SR 24.16* 3.89 <0.001 15.37 32.95

YES SR 25.45* 3.54 <0.001 17.45 33.46

OG to ATMP3 Mean difference

Standard error

Sig.a Lower Bound

Upper Bound

Pelvis NO SR -28.178* 4.51 <0.001 -38.37 -17.98

YES SR -32.01* 3.66 <0.001 -40.29 -23.73

Withers NO SR -5.83 4.21 0.200 -15.35 3.71

YES SR -29.46* 7.46 0.003 -46.32 -12.59

NO SR to YES SR Mean difference

Standard error

Sig.a Lower Bound

Upper Bound

Pelvis OG 2.09 1.17 0.107 -0.55 4.73

ATMP3 -1.75 1.48 0.267 -5.08 1.59

Withers OG 23.18* 7.13 0.010 7.04 39.32

ATMP3 -0.45 2.51 0.861 -6.14 5.23


172

Figure 6.3: Mean (±SEM) vertical displacement amplitudes (in mm) of the equine pelvis (left) and withers (right) when trotting in a straight line overground and when trotting at a water depth of mid P3 on the aqua-treadmill as informed by the Xsens© data both with and without side reins on the left (green) and right (pink) diagonal stride pair (n=10). No side reins indicated by dots. Side reins indicated by dashes. In the pelvis, vertical displacement amplitudes were significantly larger on the aqua-treadmill than overground both without a (p < 0.01) and with b (p < 0.01) side reins. In the withers, vertical displacement amplitudes were significantly larger on the aqua-treadmill with side reins c (p < 0.01).

0

10

20

30

40

50

60

70

80

90

100

110

NO side reins YES side reins NO side reins YES side reins

Overground ATM Mid P3

Dis

pla

cem

en

t (m

m)

Overground or Aqua-treadmill water at Mid P3

Pelvis

Left Right a

b

b

0

10

20

30

40

50

60

70

80

90

100

110

NO side reins YES side reins NO side reins YES side reins


Dis

pla

cem

en

t (m

m)


Withers

Left Right

c

ca

173

6.3.3 Pitch Angles Trotting Overground

The pitch of the inertial measurement units at the pelvis and withers when trotting

overground was quantified and the means plotted (Figure 6.4).


pelvis or withers, side reins and left or right diagonal pair stride on the mean pitch

angle amplitudes throughout a stride cycle of ten horses trotting overground.

There was no statistically significant three-way interaction between ‘pelvis or

withers’, ‘side reins’ or ‘left or right diagonal pair’ (F(1,9) = 0.867, p = 0.376). There

was one statistically significant two-way interaction between ‘pelvis or withers’ and

‘left or right’ (F(1,9) = 18.459, p = 0.002) requiring further analysis.

Post hoc analysis of simple main effect and pairwise comparisons showed that in

the pelvis there was an apparent asymmetry in pitch amplitudes between the left

and right diagonal pair with the right diagonal pair reaching significantly greater

pitch amplitudes both without side reins (p = 0.001), a mean (±SEM) difference of

1.82 (±0.37) degrees, and with side reins (p < 0.001), a mean (±SEM) difference of

1.52 (±0.21) degrees (Table 6.4).

Overall, the withers had lower pitch amplitudes throughout a stride cycle than the

pelvis (p = 0.039), and specifically these differences were on the right diagonal

pair both without side reins (p = 0.013), a mean (±SEM) difference of 2.69 (±0.87)

degrees and with side reins (p = 0.006), a mean (±SEM) difference of 2.43 (±0.68)

degrees (Table 6.5).

174

Table 6.4: Post hoc analysis of the significant simple main effect of left versus right on the pitch of the pelvis or withers in horses trotting overground (n=10). Pairwise comparison with Bonferroni correction shown. Mean difference shown in degrees.


Left to Right Mean

difference Standard

error Sig.a

Lower Bound

Upper Bound

Pelvis NO SR -1.82* 0.37 0.001 -2.65 -0.99

YES SR -1.52* 0.21 <0.001 -2.00 -1.03

Withers NO SR 0.07 0.35 0.851 -0.72 0.86

YES SR -0.00 0.27 0.993 -0.62 0.61

Based on estimated marginal means a Bonferroni adjustment for multiple comparisons * The mean difference is significant at the 0.05 level Table 6.5: Post hoc analysis of the significant simple main effect of pelvis versus withers on the pitch of the pelvis or withers in horses trotting overground (n=10). Pairwise comparison with Bonferroni correction shown. Mean difference shown in degrees.



difference Standard

error Sig.a

Lower Bound

Upper Bound

NO Side Reins Left 0.81 0.85 0.369 -1.12 2.73

Right 2.69* 0.87 0.013 0.73 4.65

YES Side Reins Left 0.91 0.65 0.194 -0.56 2.38

Right -2.43* 0.68 0.006 0.89 3.96


175

Figure 6.4: Mean (±SEM) pitch (in degrees) of the equine pelvis (left) and withers (right) when trotting in a straight line overground as informed by the Xsens© data both with and without side reins on the left (green) and right (pink) diagonal stride pair(n=10). No side reins indicated by dots. Side reins indicated by dashes. In the pelvis, pitch amplitudes were significantly greater on the right diagonal pair both without a (p < 0.01) and with side reins b (p < 0.01). The withers had significantly lower pitch amplitudes than the pelvis c (p < 0.05).

ab

a b

0

1

2

3

4

5

6

7

8

NO SR YES SR

OVERGROUND

Me

an A

ngl

e (°)

Pelvis

Left Right

0

1

2

3

4

5

6

7

8

NO SR YES SR

OVERGROUNDM

ean

An

gle

(°)

Withers

Left Rightc

c

176

6.3.4 Comparison between overground and ATMP3 Pitch Angles

A comparison between the overground pitch amplitudes and the amplitudes

achieved when trotting on the aqua-treadmill with just a very low water depth (mid

P3) was conducted to identify what effect, if any, a small amount of water has on

pitch amplitudes throughout a stride cycle (Figures 6.5).

A repeated measures ANOVA (2x2x2x2) was conducted. The four-way interaction

between ‘OG or ATMP3’, ‘P or W’, ‘side reins’ and ‘L or R’ was not statistically

significant (F(1,9) = 0.004, p = 0.949). There was one statistically significant three-

way interaction; ‘OG or ATMP3’, ‘P or W’, ‘L or R’ (F(1,9) = 12.711, p = 0.006)

which required further analysis.

Post hoc analyses of simple two-way interactions and simple simple main effects

with a Bonferroni correction showed that there was no significant effect of the use

of side reins on the pitch amplitudes, as side reins did not feature in the only

significant three-way interaction. The important comparison between overground

pitch amplitudes and ATMP3 pitch amplitudes showed that pitch amplitudes were

significantly higher in the pelvis when trotting overground than when trotting on the

aqua-treadmill with water at mid P3 (on the right diagonal pair only) (p = 0.002), a

mean (±SEM) difference of 2.01 (±0.45) degrees (Table 6.6). In the withers, pitch

amplitudes were significantly higher when trotting on the aqua-treadmill at a water

depth of mid P3 than they were when trotting overground on both the left diagonal

pair (p = 0.001), a mean (±SEM) difference of 1.66 (±0.36) degrees and right

diagonal pair (p < 0.001), a mean (±SEM) difference of 1.71 (±0.31) degrees

(Table 6.6).

177

Overground, pitch amplitudes were higher on the right diagonal pair than the left

diagonal pair (p = 0.001), and higher in the pelvis than withers on the right

diagonal pair (p = 0.013) as reported in Chapter 6.3.3 (Table 6.6).

Table 6.6: Post hoc analysis of the significant simple two-way interactions and simple simple main effects for the comparison in pitch amplitudes of the pelvis and withers of horses trotting overground and on an aqua-treadmill at water depth of mid P3. Pairwise comparisons with Bonferroni correction shown. Mean difference shown in degrees.



difference Standard

error Sig.a

Lower Bound

Upper Bound

OG Left 0.81 0.85 0.369 -1.12 2.73

Right 2.69* 0.87 0.013 0.73 4.65

ATMP3 Left -1.26 0.57 0.056 -2.55 0.04

Right -2.01* 0.45 0.002 0.99 3.02

OG to ATMP3 Mean

difference Standard

error Sig.a

Lower Bound

Upper Bound

Pelvis Left 0.41 0.67 0.559 -1.11 1.92

Right 2.01* 0.45 0.002 0.99 3.02

Withers Left -1.66* 0.36 0.001 -2.47 -0.84

Right -1.71* 0.31 <0.001 -2.42 -1.00

Left to Right Mean

difference Standard

error Sig.a

Lower Bound

Upper Bound

Pelvis OG -1.82* 0.37 0.001 -2.65 -0.99

ATMP3 -0.22 0.25 0.407 -0.78 0.35

Withers OG 0.07 0.35 0.851 -0.72 0.86

ATMP3 0.01 0.23 0.967 -0.52 0.54


178

Figure 6.5: Mean (±SEM) pitch amplitudes throughout a stride cycle (in degrees) of the equine pelvis (left) and withers (right) when trotting in a straight line overground and when trotting at a water depth of mid P3 on the aqua-treadmill as informed by the Xsens© data, both with and without side reins on the left (green) and right (pink) diagonal stride pair (n=10). No side reins indicated by dots. Side reins indicated by dashes. In the pelvis on the right diagonal pair, pitch amplitudes were significantly higher overground than on the aqua-treadmill a (p < 0.01). In the withers, pitch amplitudes were significantly higher on the aqua-treadmill than overground b (p < 0.01).

0

1

2

3

4

5

6

7

8


OVERGROUND Mid P3

An

gle

(°)

OG or ATMP3 and Side Rein (SR) status

Pelvis Left Right

0

1

2

3

4

5

6

7

8


OVERGROUND Mid P3

An

gle

(°)

OG or ATM and Side Rein (SR) status

Withers

Left Righta

b

179

6.3.5 Mediolateral Displacements at the pelvis, withers and poll and when

trotting overground

Mediolateral displacement amplitudes when trotting on an aqua-treadmill at

increasing water depths were discussed in Chapter 5. Here, mediolateral

displacements at the pelvis and withers were investigated when the horse is

trotting in a straight line overground, and for the first time, the poll is also included

in the analysis to determine any effect of side reins on the mediolateral

movements of the head. Means were plotted and can be seen in Figure 6.6.

A two-way repeated measures ANOVA (3x2) was conducted. There was no

significant two-way interaction between anatomical area and the use of side reins

(F(2,18) = 0.525, p = 0.494). With no significant interaction, the main effects could

be interpreted and analysed.

There was no significant main effect of anatomical location (pelvis, withers, poll)

on mediolateral displacement amplitudes (F(2,18) = 0.678, p = 0.520). No one area

appeared to have smaller or larger mediolateral displacements than another.

The main effect of side reins on mediolateral displacements proved not to be

significant (F(1,9) = 1.319, p = 0.280). It was anticipated that the use of side reins

would perhaps have had the effect of reducing mediolateral displacement

amplitudes particularly at the withers and poll but while a trend was seen (Figure

6.6) this trend was not found to be significant.

180

Figure 6.6: Mean (±SEM) mediolateral displacement amplitudes (in mm) of the equine pelvis, withers and poll when trotting in a straight line overground as informed by the Xsens© data both without side reins (orange) and with side reins (green) (n=10). There was no significant effect of side reins on mediolateral displacements of the pelvis, withers or poll when trotting overground.

0

10

20

30

40

50

No SR Yes SR No SR Yes SR No SR Yes SR

Pelvis Withers Poll

Dis

pla

cem

ent

(mm

)

Anatomical area of sensor placement

Mean Mediolateral Displacement Amplitudes when Trotting Overground

181

6.3.6 Comparison between overground and ATMP3 mediolateral

displacements (pelvis, withers and poll)

As previously stated, mediolateral displacement amplitudes on the aqua-treadmill

are reported in Chapter 5. A comparison was made between mediolateral

displacements achieved when trotting overground to trotting on the aqua-treadmill

at a low water depth of mid P3 at the pelvis, withers and poll and the means were

plotted (Figure 6.7).

A repeated measures ANOVA (2x3x2) was conducted to determine any effects of

overground or ATMP3, pelvis, withers or poll, or the use of side reins on the

mediolateral displacement amplitudes of ten horses trotting. There was no

statistically significant interaction between OG or ATMP3, pelvis withers or poll,

and the use of side reins (F(2,18) = 2.894, p = 0.081). There was a statistically

significant two-way interaction between OG or ATMP3 and pelvis, withers or poll

(F(2,18) = 7.698, p = 0.004) so further analysis was required. Post hoc analyses of

simple main effects and pairwise comparisons were performed, and six significant

effects were found.

When comparing overground trotting to trotting on the aqua-treadmill at a water

depth of mid P3 without side reins, the pelvis showed significantly greater

mediolateral displacements on the aqua-treadmill (p = 0.041), a mean (±SEM)

difference of 9.97 (±4.18) mm. This was not significant when side reins were

applied (p = 0.731). The withers showed significantly greater mediolateral

displacements on the aqua-treadmill both without and with side reins (p < 0.001), a

mean (±SEM) difference of 20.26 (±2.62) mm without side reins, and a mean

182

(±SEM) difference of 21.08 (±2.46) mm with side reins. (Table 6.8). There was no

significant difference in mediolateral displacement amplitudes in the poll from

trotting overground to trotting on the aqua-treadmill with a small amount of water

(Table 6.7).

When studying the changes in mediolateral displacements between the pelvis,

withers and poll, no significant differences were found in the overground data but

when trotting on the aqua-treadmill at a water depth of mid P3, the withers have a

significant greater mean (±SEM) mediolateral displacement than the pelvis of

15.49 (±5.15) mm when there are no side reins (p = 0.044) and when the side

reins are applied the mediolateral displacement is even greater at 21.38 (±3.91)

mm (p = 0.001) (Table 6.9). Also, with side reins, the withers have a significantly

greater mediolateral displacement than the poll of 19.47 (±5.61) mm (p = 0.021)

(Table 6.8).

Table 6.7: Post hoc analysis of the significant simple main effect of OG versus ATMP3 on mediolateral displacements of the pelvis, withers and poll of horses trotting overground and on an aqua-treadmill at water depth of mid P3. Pairwise comparisons with Bonferroni correction shown. Mean difference shown in millimetres.


OG to ATMP3 Mean

difference Standard

error Sig.a

Lower Bound

Upper Bound

Pelvis NO side reins -9.97* 4.18 0.041 -19.42 -0.52

YES side reins -0.86 2.43 0.731 -6.35 4.63

Withers NO side reins -20.26* 2.62 <0.001 -26.18 -14.35

YES side reins -21.08* 2.46 <0.001 -26.63 -15.52

Poll NO side reins -3.67 5.51 0.523 -16.13 8.80

YES side reins -0.14 3.13 0.967 -7.21 6.94


183

Table 6.8: Post hoc analysis of the significant simple main effect of anatomical location on mediolateral displacements of the pelvis, withers and poll of horses trotting overground and on an aqua-treadmill at water depth of mid P3. Pairwise comparisons with Bonferroni correction shown. Mean difference shown in millimetres.


Mean difference

Standard error

Sig.a Lower Bound

Upper Bound

OV

ERG

RO

UN

D NO

SR

Pelvis Withers -5.20 3.74 0.592 -16.16 5.76

Poll -5.12 6.29 1.000 -23.55 13.32

Withers Pelvis 5.20 3.74 0.592 -5.76 16.16

Poll 0.09 6.67 1.000 -19.47 19.64

Poll Pelvis 5.12 6.29 1.000 -13.32 23.55

Withers -0.09 6.67 1.000 -19.64 19.47

YES

SR

Pelvis Withers -1.16 3.29 1.000 -10.80 8.48

Poll -2.63 5.58 1.000 -19.00 13.74

Withers Pelvis 1.16 3.29 1.000 -8.48 10.80

Poll -1.47 4.96 1.000 -16.01 13.07

Poll Pelvis 2.63 5.58 1.000 -13.74 19.00

Withers 1.47 4.96 1.000 -13.07 16.01

ATM

P3

NO

SR

Pelvis Withers -15.49* 5.15 0.044 -30.61 -0.37

Poll 1.19 2.84 1.000 -7.13 9.52

Withers Pelvis 15.49* 5.15 0.044 0.37 30.61

Poll 16.69 5.69 0.050 -0.00 33.74

Poll Pelvis -1.19 2.84 1.000 -9.52 7.13

Withers -16.69 5.69 0.050 -33.37 0.00

YES

SR

Pelvis Withers -21.38* 3.91 0.001 -32.84 -9.91

Poll -1.91 3.65 1.000 -12.60 8.79

Withers Pelvis 21.38* 3.91 0.001 9.91 32.84

Poll 19.47* 5.61 0.021 3.01 35.93

Poll Pelvis 1.91 3.65 1.000 -8.79 12.60

Withers -19.47* 5.61 0.021 -35.93 -3.01


184

Figure 6.7: Mean (±SEM) mediolateral displacement amplitudes of the equine pelvis, withers and poll (in mm) in trot both overground and on the aqua-treadmill at a water depth of mid P3 both without side reins (orange) and with side reins (green) as informed by the Xsens© data (n=10). Mediolateral displacements were significantly larger on the aqua-treadmill than overground in the pelvis without side reins a (p < 0.05). Mediolateral displacements were significantly larger on the aqua-treadmill than overground in the withers both without b (p < 0.01) and with side reins c (p < 0.01). On the aqua-treadmill, the withers have a greater mediolateral displacement than the pelvis without side reins d (p < 0.05) and with side reins e (p < 0.01); the withers have a greater mediolateral displacement than the poll with side reins f (p < 0.05).

a

bc

ad

e

bd

cef

f

0

10

20

30

40

50

60

70

NO SR YES SR NO SR YES SR NO SR YES SR NO SR YES SR NO SR YES SR NO SR YES SR

Pelvis Withers Poll Pelvis Withers Poll


Dis

pla

cem

en

t (m

m)


Mean Mediolateral Displacment Amplitudes

185

6.3.7 Roll of the pelvis, withers and poll when trotting overground

The roll amplitudes of the inertial measurement units when trotting on an aqua-

treadmill at increasing water depths were discussed in Chapter 5. The roll at the

pelvis and withers were also investigated when the horse was trotting in a straight

line overground, and for the first time, the poll was also included in the analysis to

determine any effect of side reins on the mediolateral movements of the head.

Means were plotted and san be seen in Figure 6.8.

A repeated measures ANOVA (3x2) was conducted. There was no statistically

significant two-way interaction between pelvis, withers or poll and the use of side

reins on roll amplitudes (F(2,18) = 0.565, p = 0.578). With no significant interaction,

the main effects were interpreted and analysed.

There was no main effect of side reins on the roll amplitudes, but there was a

significant main effect of anatomical location on roll amplitudes (F(2,18) = 69.372, p

< 0.001) where pairwise comparisons showed that the poll had a significantly

lower mean (±SEM) roll amplitude than both the pelvis of 10.26 (±0.98) degrees (p

< 0.001) and the withers of 11.56 (±1.27) degrees (p < 0.001) (Table 6.9).

186

Table 6.9: Post hoc analysis of the significant main effect of anatomical location on the roll of the pelvis withers and poll of horses trotting overground (n=10). Pairwise comparison with Bonferroni correction shown. Mean difference shown in degrees.


Anatomical Location Mean

difference Standard

error Sig.a

Lower Bound

Upper Bound

Pelvis Withers -1.30 0.94 0.63 -4.06 1.46

Poll 10.26* 0.98 <0.001 7.39 13.13

Withers Pelvis 1.30 0.94 0.603 -1.46 4.06

Poll 11.56* 1.27 <0.001 7.82 15.29

Poll Pelvis -10.26* 0.98 <0.001 -13.13 -7.39

Withers -11.56* 1.27 <0.001 -15.29 -7.82


187

Figure 6.8: Mean (±SEM) mediolateral roll amplitudes (in degrees) of the equine pelvis, withers and poll when trotting in a straight line overground both without side reins (orange) and with side reins (green) as informed by the Xsens© data (n=10). There was no effect of side reins on roll amplitudes. The poll had significantly lower roll amplitudes than the pelvis a (p < 0.01) and withers b (p < 0.01).

0

5

10

15

20

No SR Yes SR No SR Yes SR No SR Yes SR

Pelvis Withers Poll

An

gle

(d

egr

ee

s)

Anatomical area of sensor placement

Mean Roll Amplitudes when Trotting Overground

a b

b

a

188

6.3.8 Comparison between overground and ATMP3 roll amplitudes

A significant effect on roll amplitudes at the poll was reported in Chapter 6.3.7

(both with and without side reins) so it was pertinent to again include the poll in the

analysis of comparing the roll amplitudes from trotting overground to the roll

amplitudes when trotting on an aqua-treadmill at a water depth of mid P3. The

means were plotted and can be seen in Figure 6.9.

A repeated measures ANOVA (2x3x2) was conducted determine any effects of

OG or ATMP3, pelvis, withers or poll, or the use of side reins on the roll

amplitudes of ten horses trotting. There was no significant three-way interaction

between ‘OG or ATMP3’, ‘pelvis, withers or poll’, and ‘side reins’ (F(2,18) = 1.431, p

= 0.265). With no significant three-way interaction, the two-way interactions were

investigated and there was one significant two-way interaction between ‘OG and

P3’ and ‘anatomical area’ (F(2,18) = 5.263, p = 0.016). Post hoc analyses of simple

main effects and pairwise comparisons were then run, and several significant

effects were found.

The comparison between overground trotting and trotting on the aqua-treadmill

showed only a significant difference in the withers. The withers showed a

statistically significant increase in roll amplitudes on the aqua-treadmill at a water

depth of mid P3 compared with overground, both without side reins (p < 0.001), a

mean (±SEM) difference of 5.77 (±1.07) degrees, and when side reins were added

(p = 0.038), a mean (±SEM) difference of 6.99 (±2.87) degrees. There were no

other significant differences in roll amplitudes between trotting overground and

trotting on the aqua-treadmill at a water depth of mid P3 (Table 6.10).

189

When trotting overground, the poll had significantly lower roll amplitudes than the

pelvis and withers both with and without side reins (p < 0.001) (as reported in

chapter 6.3.7) but there was no significant difference in the roll amplitudes

achieved between the pelvis or withers (Table 6.11).

When trotting on the aqua-treadmill at a water depth of mid P3 without side reins,

the withers had significantly greater roll amplitudes than the pelvis (p = 0.010), a

mean (±SEM) difference of 6.47 (±1.64) degrees, and the poll (p < 0.001), a mean

(±SEM) difference of 14.80 (±2.19) degrees, and the poll had significantly lower

roll amplitudes than the pelvis (p = 0.001), a mean (±SEM) difference of 8.33

(±1.57) degrees. When side reins were added, the withers still had greater roll

amplitudes than the poll (p < 0.001), a mean (±SEM) difference of 19.05 (±2.80)

degrees, but not the pelvis (p = 0.112). The poll still had significantly lower roll

amplitudes than the pelvis (p < 0.001), a mean (±SEM) difference of 11.71 (±1.12)

degrees (Table 6.11).

Table 6.10: Post hoc analysis of the significant simple main effect of overground versus ATMP3 on roll amplitudes of the pelvis, withers and poll of horses trotting overground and on an aqua-treadmill at water depth of mid P3. Pairwise comparison with Bonferroni correction shown. Mean difference shown in degrees.


OG to ATMP3 Mean

difference Standard

error Sig.a

Lower Bound

Upper Bound

Pelvis NO side reins -0.94 0.90 0.323 -2.97 1.09

YES side reins -0.62 0.58 0.317 -1.93 0.70

Withers NO side reins -5.77* 1.07 <0.001 -8.18 -3.36

YES side reins -6.99* 2.87 0.038 -13.49 -0.49

Poll NO side reins -2.73 2.27 0.260 -7.87 2.41

YES side reins 0.71 0.50 0.191 -0.43 1.85


190

Table 6.11: Post hoc analysis of the significant simple main effect of anatomical location on roll amplitudes of the pelvis, withers and poll of horses trotting overground and on an aqua-treadmill at water depth of mid P3. Pairwise comparison with Bonferroni correction shown. Mean difference shown in degrees.


Mean difference

Standard error

Sig.a Lower Bound

Upper Bound

OV

ERG

RO

UN

D NO

SR

Pelvis Withers -1.64 1.14 0.553 -5.00 1.71

Poll 10.12* 1.01 <0.001 7.16 13.08

Withers Pelvis 1.64 1.14 0.553 -1.71 5.00

Poll 11.77* 1.24 <0.001 8.12 15.41

Poll Pelvis -10.12* 1.01 <0.001 -13.08 -7.16

Withers -11.77 1.24 <0.001 -15.41 -8.12

YES

SR

Pelvis Withers -0.96 0.80 0.778 -3.29 1.38

Poll 10.39* 1.10 <0.001 7.17 13.60

Withers Pelvis 0.96 0.80 0.778 -1.38 3.29

Poll 11.35* 1.36 <0.001 7.35 15.35

Poll Pelvis -10.39* 1.10 <0.001 -13.60 -7.17

Withers -11.35* 1.36 <0.001 -15.35 -7.35

ATM

P3

NO

SR

Pelvis Withers -6.47* 1.64 0.010 -11.28 -1.67

Poll 8.33* 1.57 0.001 3.72 12.95

Withers Pelvis 6.47* 1.64 0.010 1.67 11.28

Poll 14.80* 2.19 <0.001 8.38 21.22

Poll Pelvis -8.33* 1.57 0.001 -12.95 -3.72

Withers -14.80 2.19 <0.001 -21.22 -8.38

YES

SR

Pelvis Withers -7.34 3.01 0.112 -16.58 1.49

Poll 11.71* 1.12 <0.001 8.43 15.00

Withers Pelvis 7.34 3.01 0.112 -1.49 16.16

Poll 19.05* 2.80 <0.001 10.83 27.27

Poll Pelvis -11.71* 1.12 <0.001 -15.00 -8.43

Withers -19.05* 2.80 <0.001 -27.27 -10.83


191

Figure 6.9: Mean (±SEM) roll amplitudes of the equine pelvis, withers and poll (in degrees) in trot both overground and on the aqua-treadmill at a water depth of mid P3 both without side reins (orange) and with side reins (green) as informed by the Xsens© data (n=10). Roll amplitudes were significantly larger on the aqua-treadmill than overground in the withers both without a (p < 0.01) and with b (p < 0.05) side reins. On the aqua-treadmill, the poll had significantly lower roll amplitudes than the pelvis c (p < 0.01) and withers d (p < 0.01). On the aqua-treadmill without side reins, the withers had larger roll amplitudes than the pelvis e (p < 0.01) and the poll f (p < 0.01), and the poll had significantly lower roll amplitudes than the pelvis g (p < 0.01). On the aqua-treadmill with side reins, the withers had larger roll amplitudes than the poll h (p < 0.01), and the poll had significantly lower roll amplitudes than the pelvis i (p < 0.01)

ab

eg i

aef

bh

fg

hi

0

10

20

30

NO SR YES SR NO SR YES SR NO SR YES SR NO SR YES SR NO SR YES SR NO SR YES SR

Pelvis Withers Poll Pelvis Withers Poll


An

gle

(d

egr

ee

s)


Mean Roll Amplitudes Overground and Trotting on the Aqua-treadmill

c d

c

d

192

6.4 Discussion

6.4.1 Overground vertical displacements

There were many elements to this study, but the key areas for discussion begin

with the information gleaned from trotting horses overground.

This study reported overground vertical pelvic displacement amplitudes in the trot

of 66.66 ± 8.76 mm (mean ± 1 SD) with a range of 49.86 to 83.15 mm which

concurs with previous studies of amplitudes of around 60mm in non-lame horses

using the Xsens© system for data collection (Church et al., 2009). Overground

withers vertical displacements were quantified as 67.71 ± 8.94 (mean ± 1 SD) with

a range of 51.97 to 86.84 mm, which agrees with an earlier study where horses

were trotted overground and withers vertical displacements were reported as a

mean of 66.0 ± 12.0 mm (mean ± 1 SD) (Buchner et al., 1994a). This agreement

in vertical displacement amplitudes gives confidence in the data collection and

handling techniques employed in this study and so gives confidence that any

effect of side reins is notable.

This study reported a significant effect of the use of side reins decreasing vertical

displacement amplitudes in the withers. This effect was not seen in the pelvis.

This suggests perhaps, that side reins have a stabilising effect on the front end of

the horse, perhaps controlling erratic head movements by keeping the head and

neck straighter and lowering the poll, thereby stabilising the movement of the

withers. It is possible that more energy is exerted into forward thrust rather than

vertical displacements and stride length may therefore become longer with the use

193

of side reins. However, stride length was not measured in this study, but a

simultaneous study investigating vertical displacements and stride parameters

would be interesting to further quantify locomotory changes overground with the

use of side reins. There appear to be no reports in the literature discussing the

effect of side reins or training aids on stride parameters, only the effect of different

length side reins on the resulting rein tensions (Clayton et al., 2011) and the effect

of different training aids on EMG activity in the longissimus dorsi (Cottriall et al.,

2009).

6.4.2 Overground mediolateral displacements

Literature has previously reported mediolateral displacements at T6 of around 34

mm but with a range of 16 to 37 mm throughout different locations on the back

decreasing from cranial to caudal and then increasing in the sacral region with the

1st lumbar vertebrae recording mediolateral displacements of around 17mm

(Warner et al., 2010). The current study reports a larger mean mediolateral

displacement in the pelvis of 36.03 ± 6.63 mm (mean ± 1 SD) with a range of

24.86 to 48.28 mm, and in the withers; 41.24 ± 11.11 mm (mean ± 1 SD) with a

range of 26.01 to 55.85 mm but does exhibit the same trend as the previous study

(Warner et al., 2010) with greater mediolateral displacements shown at the withers

than at the pelvis. This pattern of lateral displacement amplitudes corresponds to

the body centre of mass which is most similar to the movement at T13 (Buchner et

al., 2000) and the further away from this, the greater the lateral movement

(Buchner et al., 2000; Warner et al., 2010).

194

Mediolateral displacements appear to be less studied in the wider scientific

literature possibly due to vertical displacements most often being studied with

regards to lameness, and symmetry indices determined from the asymmetries in

vertical displacements between the tuber coxae and sacrum. With regard to the

mediolateral displacement amplitudes when trotting overground, side reins

appeared to reduce mediolateral displacements in the withers, and again, perhaps

stabilise the front end of the horse. Although this trend was not found to be

significant, a supplementary investigation utilising a greater cohort of horses may

develop this trend further. It is suggested, however, that mediolateral

displacements in the withers are also subject to perhaps a greater variation in skin

rolling over the withers as skin is known to move over bony landmarks, and up to

as much as 12cm in proximal parts of the limb (Clayton and Schamhardt, 2007).

Plus, there can be difficulty in the fixation of the inertial measurement unit over the

withers hence a specific withers mount is required (Pfau et al., 2005).

6.4.3 Overground roll

The atlantooccipital joint has been documented as able to rotate through a mean

of 27 degrees in cadaver studies (Clayton and Townsend, 1989) whereas the

current study reports a mean roll at the poll in trotting horses of 5.90 ± 1.84

degrees (mean ± 1 SD) with a range of 3.53 to 9.50 degrees. Roll amplitudes

overground in the present study were reported in the pelvis as 16.02 ± 2.45

degrees (mean ± 1 SD) with a range of 11.55 to 19.41 degrees, and in the withers

as 17.67 ± 3.72 degrees (mean ± 1 SD) with a range of 12.56 to 25.53 degrees.

195

Haussler et al. (2001) reported axial rotation in segmental vertebrae in the trot as a

mean of 4.4 ± 0.8 through the region of L6 to S1 which was as close anatomically

as could be found in current literature that could be compared to the roll of the

tuber sacrale which was measured in the current study. Haussler et al. (2001)

reported figures a lot less than the mean 16.02 ± 2.45 degrees reported in the

pelvis in the current study but this could be due to the fact that the horses were

trotting on a dry treadmill not overground, and actually the 2001 study only

reported the mean of three horses. On a high-speed treadmill, pelvic roll was

reported to be 6.9 ± 2.1 degrees at trot with a relaxed head and neck position

(Gomez Alvarez et al., 2006) which is still not as high as the values seen

overground in the present study. There may simply be a significant change in

pelvic roll in treadmill locomotion compared to overground which has not yet been

quantified in current literature and so is a potential area for further research.

However, Pfau et al. (2005) also reports lesser roll amplitudes at the withers of

around 10 degrees recorded from inertial measurement units, compared to 17.67

± 3.72 degrees (mean ± 1 SD) in the present study. Again, this may simply be due

to differences in dry treadmill and overground locomotion which warrants further

investigation.

In the present study, the roll of the inertial measurement units in horses trotting

overground, showed the poll had significantly less roll than both the pelvis and

withers. This is to be expected due to the anatomical structure of the

atlantooccipital joint where there is only limited rotation permissible at this joint.

Side reins had no effect of increasing or decreasing this roll, and had no effect on

the roll in the pelvis or withers either. The lack of difference in roll amplitudes

196

between the pelvis and withers suggests perhaps that the horses in this study

were actually well balanced with no apparent lameness or asymmetries which

corroborates with the Vet that deemed them non-lame and suitable to take part in

the study.

6.4.4 Effect of water on vertical displacements

Investigating the effect of a small amount of water on vertical and mediolateral

displacements had interesting findings. Firstly, it seems that a small amount of

water does indeed induce a significant effect in vertical displacements particularly

at the pelvis and without the use of side reins. However, this result could simply

be due to the differences in treadmill locomotion compared to overground, rather

than specifically due to the water but a dry treadmill was not included in the current

study. Previous literature has shown that vertical displacements of the withers

were comparable from overground trotting on asphalt to trotting on a dry treadmill

(Buchner et al., 1994a). The speed of the aqua-treadmill was matched to the

speed the horses trotted in hand overground which was an important factor, as

kinematic stride variables have been reported to be velocity dependent (Leach and

Drevemo, 1991; Buchner et al., 1994a). The pelvis exhibited significantly greater

vertical displacements on the aqua-treadmill with water at a depth of mid P3 than it

did when trotting overground. This effect was not seen in the withers which

suggests that the withers can compensate the water depth by being able to bend

at the carpus, whereas the stay and reciprocal apparatus in the hindlimb creates

significant flexion throughout the whole limb increasing compression and therefore

increasing vertical lift. It is suggested therefore, that a very low depth of water

197

induces a large response in hindlimb flexions which may in turn create a greater

workload for the horse. This potentially corresponds to the effects of tactile

stimulation of the pastern and coronary band where significant increases in peak

flexion of the stifle, tarsal, MTP and DIP joints throughout the swing phase in trot

when a lightweight tactile stimulator (55 g) consisting of a 1 cm wide braided nylon

strap with 7 double strands of lightweight brass chain, 6 cm in length, were

attached loosely around the hind pasterns (Clayton et al., 2008; 2010). It is

suggested that the effect of tactile stimulation would not diminish with increasing

water depth but the extra stimulus effects of the resistance and weight of the water

would perhaps encourage larger muscle development while maintaining this

increased range of motion.

6.4.5 Effect of side reins

It is interesting that when trotting overground, side reins had a significant effect of

reducing vertical displacements in the withers, but when trotting on the aqua-

treadmill at a water depth of mid P3 this effect was not seen. This could indicate

that there is a stabilising effect of water keeping both the hind and front end of the

horse more stable. It could be in fact, that overground, the side reins have the

effect of controlling the head, but perhaps also making the horse more ‘downhill’.

The water could serve to counteract this effect by controlling and straightening the

head and neck but without creating the downhill effect that was seen in the

significantly lower vertical displacements overground.

198

6.4.6 Pitch

There was an interesting anomaly seen with regards to the pitch amplitudes

achieved by the inertial measurement units when trotting overground; the right

diagonal pair had significant greater pitch amplitudes throughout a stride cycle

than the left diagonal pair. Considering this effect was not seen when trotting on

the aqua-treadmill, it is possible that there may have been an effect of handler, or

more likely, an effect of the incline of the tarmac road used to trot the horses. A

significant difference in pitch range between level and incline galloping on a

treadmill has been reported (Parsons et al., 2008). Side reins had no significant

effect of increasing or decreasing pitch amplitudes. The pelvis displayed greater

pitch amplitudes when trotting overground, but the reason for the statistical

significance here could be due to the right diagonal pair anomaly (which is likely

due to the presence of a handler). Looking at the left diagonal pair, there was no

greater pitch amplitudes achieved overground in the pelvis.

The most interesting result was the significant greater pitch amplitudes seen in the

withers when trotting on the ATMP3 compared to when trotting overground. Even

with no effect of side reins, this perhaps suggests that the horse has to ‘pitch up’

and therefore perhaps ‘jump’ over the water at the front end. It has been

suggested in this thesis that the front end of the horse may compensate the water

depth by being able to bend at the carpus, but the increase in pitch seen here also

suggests perhaps that there is hyperextension of the neck from the withers to

create a jump at the front end.

199

6.4.7 Effect of water on mediolateral displacements

No significant effects were seen in the mediolateral displacements overground, but

when trotting in a small amount of water on the aqua-treadmill without side reins,

both the pelvis and withers had significant increases in mediolateral

displacements. This perhaps suggests that the increased effort that results in the

increased vertical displacements from an increased compression of the limb, has

the effect of tipping the horse side to side thus increasing mediolateral

displacements. The results of the investigation into the roll would perhaps confirm

this in the withers. The roll of the withers was significantly increased in the aqua-

treadmill compared to overground, but again, the potential for skin movement over

the withers may be the contributing factor here. Side reins also had the effect of

reducing mediolateral displacements on the ATM at mid P3; indicating that

perhaps side reins have the effect of stabilising the pelvis when working in water.

It is appreciated that a direct comparison between a dry treadmill belt and a

treadmill belt with a small amount of water would perhaps be preferred, but as

aqua-treadmill belts do not like to run ‘dry’ then this was not possible on this

occasion. Also, it is suggested that the parameters measured in this study may

not be directly affected by locomotion on a belt. Vertical and mediolateral

displacements are perhaps unlikely to change whether moving on a motorised belt

or overground as previously discussed. Also, this study sought to investigate the

effect of just a small amount of water on the measured parameters which it was

successful in doing. Indeed, it appears that only a small amount of water is

required to induce a significant effect in both vertical and mediolateral

displacements.

200

6.5 Conclusion

This chapter has identified that there is a clear effect of a very low depth of water

on the biomechanical parameters of vertical and mediolateral displacements, pitch

and roll, measured at the pelvis, withers and poll. Side reins decrease vertical

displacement amplitudes in the withers overground by controlling erratic

movements of the head and neck but perhaps make the horse more downhill.

Trotting on an aqua-treadmill with only a small amount of water counteracts this

downhill effect. Side reins also supported the stability of the hindquarters when

working through water preventing excessive mediolateral movements.

Trotting through a very low water depth generated significantly greater vertical

displacements in the pelvis than when trotting overground and significantly greater

vertical displacements than the withers suggesting that the withers can

compensate the water depth by being able to flex at the carpus, but also having an

increased pitch at the withers suggests more extension to create a jump up over

the water. A very low water depth is therefore suggestive of a significantly greater

workload for the horse in both the front end and hindquarters than overground

trotting. The significantly increased vertical displacements in both the pelvis and

withers when trotting through a small amount of water compared to overground is

also indicative of greater limb compression and a greater workload. It is therefore

suggested that the aqua-treadmill is beneficial at increasing the workload for the

horse that will have a corresponding effect of increasing muscle mass, strength

and condition, but without detrimental effects to cranial-caudal or mediolateral

symmetry patterns.

201

CHAPTER 7 Discussion and Conclusions

7.0 Introduction

Equine aqua-treadmills have become increasingly popular in practice and scientific

literature is growing. There is however, still relatively little research documenting

how horses move during this method of locomotion and what deviations exist from

overground locomotion. With more emphasis and investment in recent years in

the competition horse to harness any opportunity to prevent injury, give horses an

extra edge in their training, to rehabilitate them from injury more quickly and

ultimately promote and prolong their competitive career, riders, trainers and

Veterinary Surgeons are anecdotally reporting numerous benefits of aqua-

treadmill exercise. The overarching aims of this study were to quantify equine

kinematics during aqua-treadmill exercise, with specific focus on the axial skeleton

during trot.

There were four main objectives for this study; 1) Determine the impact of water

depth on vertical pelvis and wither displacement while trotting on an aqua-treadmill

(Chapter 3); 2) Determine the effect of side reins on vertical displacements of the

pelvis and withers when trotting on an aqua-treadmill at increasing water depths

(Chapter 4); 3) Determine the impact of water depth on mediolateral pelvis and

wither displacements when trotting on an aqua-treadmill (Chapter 5); and 4)

Determine the difference a small amount of water makes to pelvis and wither

displacements when trotting on the aqua-treadmill compared to overground trotting

(Chapter 6). In order to address these objectives, novel data were successfully

collected, processed and analysed. This data provides an insight, developing a

202

better understanding of equine locomotion in each specific scenario, proposing

likely adaptation strategies used by the horses. These data and observations can

be used to inform current practices and better tailor training and rehabilitation

programmes to the individual needs of the horse from the leisure horse to the elite

competition horse.

This chapter provides a summary of the findings presented in previous chapters,

tying all the relevant information together to start and build one big picture. This

combined information can then be applied to the industry to demonstrate the

significance and relationship to current scientific knowledge and the scope of

future developments. Furthermore, it will highlight the limitations of current

knowledge in conjunction with areas that have raised further questions requiring

further investigation.

7.1 Effects of water depth

Water depth was found to have a significant impact on equine kinematics

indicating changes in locomotor technique when trotting through water of varied

depth. Increasing water depth up to the carpus resulted in increased vertical

displacement amplitudes at both the withers and pelvis. These increases in

vertical displacement suggest the horse had to alter their locomotor pattern

working differently to lift themselves up and out of the water. A previous study on

heart rates by Lindner et al. (2003) found no significant effect of water depth on

heart rates in either walk or trot which would disagree with the suggestion that the

horse is having to work harder in increased water depths, although a later study by

203

Nankervis et al. (2008a) concluded that heart rate is not a reliable indicator of work

done during partial submersion with a range of different depths, as water

temperature has a greater effect on heart rates than depths. The greater vertical

displacements achieved at deeper depths may also be due in part to the buoyancy

of the water, and of course the speed of the trot where optimum energy preserving

system of the spring-mass model is affected by velocity (Blickhan, 1989; Farley et

al., 1993). The pelvis had significantly larger vertical displacements than the

withers at all water depths. This was an interesting and unexpected finding and

was not seen in the overground study indicating that water depth has a larger

impact on the hindlimbs than the forelimbs. This difference between withers and

pelvis is likely due directly to the anatomy of the horse. The stay and reciprocal

apparatus of the hindlimb may be affected by the water depth to a greater extent

than the forelimbs, with the forelimbs perhaps being able to compensate for the

increasing water depths by flexing at the carpus providing an ease of movement

up and over the water. This would provide optimal conditions from a training

perspective, where horses are encouraged to work from behind more and lighten

the forehand as defined by the Federation Equestre Internationale (FEI) Rulebook

for Dressage 2017 (Anon, 2017). It is unknown from the current study which

muscles are recruited to create these locomotory changes but it could be expected

that undertaking desirable movements such as these would likely activate and

develop the same muscles as would be required to provide such movements

overground under saddle thus improving performance. To further consolidate this

theory and demonstrate this optimal movement and expected weight distribution,

establishing that the muscles of the hindlimb are working harder than the forelimb,

a further study investigating muscle activity perhaps by electromyography would

be required despite the clear changes in the mechanics of the limb. EMG activity

204

of the back muscles and hind limbs of horses has successfully been reported

during trotting (Robert et al., 2000; Robert et al., 2001a, 2001b; Robert et al.,

2002; Licka et al., 2004) with the most recent study establishing that during take-

off of a hind limb, EMG activity of the longissimus dorsi muscle ipsilateral to that

limb is higher than EMG activity of the contralateral longissimus dorsi muscle,

which was attributed to the propulsive force generated by the hind limb, and

maximum EMG activity was found to precede the maximum vertical excursion,

which could be interpreted as pre-emptive tension (Licka, 2004). Increasing speed

has also been found to effect EMG activity of muscles of the hindlimb during trot

(Robert et al., 2002) where changes in speed could be related to the mechanical

function of the muscles involved and greater trotting speeds are associated with

greater vertical ground reaction forces (Barr et al., 1995) and require increased

muscle activity. A more recent study investigated work done by both fore and

hindlimbs using ground reaction forces and found that from inverse dynamic

analysis, the forelimb did not appear to do any work (i.e. it neither absorbing or

generating energy), while the hindlimb did positive work at all speeds (Dutto et al.,

2006). This potentially concurs with the findings of the present study with the hind

limb apparently working harder than the forelimb to create the greater vertical

displacements seen when trotting on the aqua-treadmill. It would be interesting to

also measure vertical ground reaction forces when trotting on an aqua-treadmill to

establish if the buoyancy of the water perhaps reduces ground reaction forces, yet

increases vertical displacements.

In the forelimb, it is the muscles of the neck that are often studied as dynamic

aspects of asymmetric head and neck movements are associated with lameness

compensation. Electromyography activity of the splenius muscle at walk and trot

205

has been reported using surface EMG (Robert et al., 2001a, b; Robert et al., 2002;

Zsoldos et al., 2010) and needle EMG (Tokuriki et al., 1999), and it has been

reported that the splenius muscle reaches maximum activity at the beginning of

the forelimb stance phases in trot, indicating functional stabilisation against flexion

of the head and neck (Zsoldos et al., 2010). The one study that has investigated

EMG activity of the forelimb muscles during aqua-treadmill exercise established

that the extensor digitorum communis showed the most intense EMG activity at

both walk and trot in the aqua-treadmill indicating that this muscle probably aided

in protracting the distal limb against the resistance of the water, as the water was

deep at a depth of 1.2 metres (Tokuriki et al., 1999). This study concluded that

trotting in the aqua-treadmill may provide less intensive training for some forelimb

muscles than walking but the depth of the water is important to note here, and

frustratingly, the study did not report any results of the EMG activity of hindlimb

muscles.

It would also be interesting to further investigate these kinematic changes by

examining flexion and extension throughout the limbs, in particular, changes in

joint angles at each discrete joint as well as the limb as a whole. Furthermore,

studying the changes at water depths deeper than that of mid carpus would allow

investigation in to how much more vertical displacement can physically be

achieved and therefore perhaps the maximum capacities of flexion and extension

in the limbs. Linking these studies with further studies into ground reaction forces,

EMG activity and also perhaps heart rates when trotting on an aqua-treadmill,

would further establish the potential work load associated with increasing water

depths.

206

When investigating the absolute positions of the pelvis and withers, the effect of

water depth was again significant with both pelvis and withers exhibiting lower

starting points of the stride during stance, suggesting an increase in both hindlimb

and forelimb compression. Increased limb compression could be determined

through close analysis of individual joint angles and may have been employed to

increase the ease of push off to compensate for the increasing depth of the water

through elastic recoil of the tendons (Dimery et al., 1986; Blickhan, 1989;

Biewener, 1998; Farley et al., 1993). The pelvis started the stride significantly

lower than the withers suggesting that the hindquarters have an increased

engagement by perhaps compressing further to work harder to create the larger

displacements and therefore increasing range of motion through a greater number

of effective joints compressing under load. Forelimbs can flex at the shoulder,

elbow and fetlock during stance, the hindlimbs can flex at the hip, stifle, hock and

fetlock meaning there is effectively an increased ability to compress the hindlimbs

underload. However, investigation of the pitch showed no corroboration with this

theory – if the pelvis was significantly ‘tucking under’ then it would be expected

that the pitch amplitude of the IMU on the tuber sacrale would highlight this,

however this was not found to be true during the current study. In contrast to the

minimum absolute vertical position and associate increased hindlimb compression,

maximum vertical position during the stride was found to significantly increase with

water depth but no significant difference was shown between the pelvis and the

withers further supporting the theory of increased hindlimb compression to allow

increased push off, aided by elastic recoil to reduce the energy cost of locomotion.

In essence the hindlimbs drop slightly lower but then realign with the forelimbs

during the aerial phase. Further investigation into limb flexions and extensions and

associated EMG activity would provide further insight into the mechanisms utilised,

207

further validating the use of the aqua-treadmill as an optimal training tool for

improving muscular strength and musculoskeletal condition in horses.

Water depth was not found to impact on the movement symmetry of the horse in

the pelvis. This further supports the application of aqua-treadmill exercise during

training and rehabilitation as it suggests that horses can continue to work harder

(as demonstrated by increased vertical displacement amplitude and increased

hindlimb compression) in the deeper water without detrimental effects to their way

of going. In contrast the withers started to show an asymmetry during trot at mid

carpus. Perhaps this highlights the physical limitations of the forelimb adaptations,

as the ability to bend at the carpus to compensate for water depth reaches its limit

and the horses then have to either jump up over the water creating some sway or

drag the distal limb through the water. Dragging the distal limb through the water

may in turn highlight muscle weakness or asymmetry as there are no energy

saving mechanisms that can be recruited. The magnitude of this asymmetric sway

may be linked to each horse’s individual strength and balance. Stronger more

balanced horses may produce less asymmetry at the withers as the water gets

deeper. This is an area for further longitudinal investigation, to determine whether

training horses on the aqua-treadmill is capable of subsequent improvement in

balance and movement symmetry.

These differences in movement symmetry between the withers and pelvis at

increased water depth are further supported by investigations into mediolateral

displacements. No significant effect of water depth has been shown, however at

the lower depths of mid P3 and mid Fetlock, the withers had greater mediolateral

displacements than the pelvis. This could be linked to and supported by a

208

published study investigating walk, where mediolateral displacements did not

decrease until water depth reached the level of the elbow and shoulder (Mooij et

al., 2013), however, it is difficult to directly compare the locomotion patterns of

walk and trot although fundamental principles associated with increasing water

depth could be expected to have a universal effect. It appeared that mediolateral

displacements had a trend of increasing at increasing water depths in the pelvis

and decreasing at increasing water depths in the withers, to the point that at the

deepest depth studied of mid carpus, the mediolateral displacement amplitudes

were comparable. Likewise, there was no effect of water depth found on

mediolateral flexions. This suggests that there is stability in the vertebral column

during trot throughout all depths of water. A further investigation into the flexion-

extension of the individual vertebrae in trot (not just the three points used in this

study; withers-T13-pelvis) would be useful to add to this study and make

comparisons with previously published data in the walk (Nankervis et al., 2016).

From a training perspective, the trends found in this investigation suggest there

are not any specific advantages or disadvantages of increased water depth on

mediolateral capabilities

Water depth had no effect on roll amplitudes but the withers did show overall

greater roll amplitudes than the pelvis which is most likely due to the superficial

sensor attachment and associated movement of skin over this area (Clayton and

Schamhardt, 2007). Overground, however, there was no significant difference

between the pelvis and withers which suggests less effect of skin movement, but a

greater effect of water. The deeper water may be creating a change in locomotory

pattern in the front end of the horse, causing the horse to really rock from side to

side. This rocking may be linked temporally with each stance phase where the

209

limb is compressing more to create the greater vertical lift induced by the presence

of water. It is expected that roll amplitudes are not likely to show large variation in

the trot, but they have been shown to increase with increasing water depth in the

walk as the increased flexion in the hindlimb creates a larger axial rotation (Mooij

et al., 2013).

7.1.1 Overground comparisons

The overground work in this study reported pelvic vertical displacements of 66.66

± 8.76 mm (mean ± 1 SD) which concurs with previous studies of amplitudes of

around 60 mm in non-lame horses (Church et al., 2009). Overground withers

vertical displacements were reported as 67.71 ± 8.94 (mean ± 1 SD) which agrees

with an earlier study that reported withers vertical displacements as a mean of

66.0 ± 12.0 mm (mean ± 1 SD) (Buchner et al., 1994a). This enabled confident

comparisons to be made to the work completed in the aqua-treadmill. A small

amount of water (water depth at mid P3) had a significant effect on the vertical

displacement amplitudes in the pelvis which further supports the theory of tactile

stimulation in the hind limbs (Clayton et al., 2008; Clayton et al., 2010) where the

water in the present study acts as a very effective tactile stimulator without being

of sufficient depth to inhibit the free movement during swing phase of the stride.

During trot, horses reduce the amount of work required by the muscles through an

exchange of potential and kinetic energy throughout the trajectory of their body’s

centre of mass. Energy exchange is achieved through the storage and recovery of

elastic strain energy in the long tendons and ligaments of the limb (Dimery et al.,

1986; Blickhan, 1989; Biewener, 1998; Farley et al., 1993). In the present study,

210

only a small amount of water was required to induce significantly greater vertical

displacements in the pelvis, and deeper water depths produced significantly larger

vertical displacements at each increasing depth, which were maintained

throughout each trial. It is interesting to note here, the differing information from

previous studies on the effect of water depth on heart rates that have evidenced

that water depth does not have a significant effect on heart rates in walk or trot

(Lindner et al., 2003; Scott et al., 2010) indicating that the increased vertical

displacement shown here is likely a result of potential and kinetic energy exchange

thus limited extra energy requirement. Previous studies on heart rates during

walking and trotting in different depths of water have been inconclusive in terms of

determining workload for the horse but in deeper water there is the element of the

limb being ‘pulled’ or ‘dragged’ through the water which is likely to increase the

energy requirement and work carried out by the muscles. The additional vertical

displacement at low water levels is likely therefore induced by the tactile

stimulation effect of the water and achieved through energy exchange in the

tendons. The significantly larger vertical displacements were maintained

throughout each trial, with no evidence of fatigue, which suggests that also greater

elastic storage was maintained throughout each of the trot trials on the aqua-

treadmill but that this greater elastic storage was initiated by the presence of water

and then amplified with the increasing depth. This does suggest that, along with

the tactile stimulation effect of the water and the increased weight of the deeper

water, it would reason that there was an increased activation of muscular

structures also, therefore deducing that simply water, and then deeper water

provides an increased work load to the musculoskeletal system and would

therefore have a conditioning and fittening effect on these systems. This is a

considerable area for further research which should include work to be carried out

211

on respiratory gasses along with studies on EMG activity to identify muscle

activation and potentially also to measure strain in the tendons which could

potentially be achieved by implanting a strain gauge. This work would be more

invasive but more informative providing a whole picture of achieved workload of

different water depths.

Mediolateral displacements in the pelvis and withers were significantly larger when

trotting on the aqua-treadmill in a small amount of water than when trotting

overground, but increases in water depth made no further significant changes. It

is likely that the exertion resulting in increased vertical displacement and

associated limb compression on the aqua-treadmill is responsible for the greater

mediolateral displacements seen. This theory is further supported by the

increased roll amplitude of the withers during aqua-treadmill exercise compared to

overground. Despite previous studies concluding that three, five-minute sessions

are sufficient to stabilise stride kinematics on a treadmill (Buchner et al., 1994b) an

alternative explanation would be that the increased mediolateral displacements

shown here were indicative of changes to the balance of the horse on the treadmill

belt compared to overground. The horses used in the present study were all

habituated to treadmill exercise, but it would be interesting to carry out a

longitudinal investigation to determine whether repeated aqua-treadmill use

improved balance and reduced mediolateral displacements.

212

7.2 Effects of side reins

Despite their common application within equine training, when exercising on the

aqua-treadmill, side reins had no significant effect on the vertical displacement

amplitudes achieved by either the pelvis or withers and side reins had no

significant effect on pitch amplitudes or displacement symmetry, leading to the

deduction that side reins may be a useful tool in an aqua-treadmill to maintain a

straight and correct head and neck position and perhaps to help sharper horses

concentrate in their work, but without any detrimental effects to movement patterns

and symmetry. It could be expected that alternative muscles were engaged for

comparative vertical displacement amplitudes to be achieved both with and

without the side reins. Previous studies have concluded that when working

horses, lowering the neck and working the horse downwards and forwards

increases the movement of the back and strengthens the back muscles (Denoix et

al., 2001) and that the hindlimbs become more engaged when the horse has a low

head position where the poll is at the same level or lower than the withers and with

a wider head/neck angle (Roepstorff et al., 2002). It is also likely that in this long

and low position the ventral muscles such as the abdominals and pectorals along

with the deeper muscles of the back become more engaged when exercising

within an aqua-treadmill as the horse must control their balance and rhythm on the

treadmill belt and work to lift themselves up over the water. This promotes

development of the desired muscles required for impulsion, engagement and

ultimately performance. Other studies have also identified that head and neck

position have an effect on stride length and the kinematics of the back (Faber et

al., 2002; Johnston et al., 2002; Rhodin et al., 2005; Gomez Alvarez et al., 2006;

Rhodin et al., 2009) but stride kinematic variables were not measured in the

213

current study so it is impossible to know if side reins had the same effects as

previously seen when exercising through water. Side reins may also be modified

to affect or target specific muscle groups in the neck and back but further studies

are required, probably using EMG, to determine which muscles are recruited with

the use of side reins, and how changes in length and positioning of the side reins,

along with changes in water depth affect muscle engagements.

It was suggested that side reins may have had a stabilizing effect on the side to

side sway seen in the withers at lower water depths, and although a reduction in

mediolateral displacement amplitudes was seen with the addition of side reins at

the lower water depths this trend was not found to be significant. This would

suggest that again, side reins have no detrimental effect to exercising horses in

water but may in fact provide beneficial, stabilising proprioceptive feedback in the

form of a contact, encouraging horses to work into the contact as they would under

saddle. Interestingly, side reins had the effect of reducing vertical displacements

in the withers when trotting overground, to the point that they were significantly

lower than that of the pelvis, but without side reins, the pelvis and withers had

comparable vertical displacements. This suggests that the side reins had the

effect of keeping the head and neck more still and potentially rendering the horse

‘downhill’. The effect of water on the aqua-treadmill could therefore serve to

counteract this ‘downhill’ effect in the withers by controlling and straightening the

head and neck but without creating the downhill effect that was seen in the

significantly lower vertical displacements overground. However, here it is pertinent

to acknowledge the importance of speed, where on the aqua-treadmill, speeds

were able to be kept constant, but overground, the horses may have been

accelerating or decelerating or losing impulsion. Although as far as possible, the

214

overground speeds were matched to those on the aqua-treadmill, but further

investigation into the impact of side reins on trot speed would also be useful. The

results of this study essentially suggest it may be more beneficial to use side reins

in an aqua-treadmill than overground. However, without quantification of the rest

of the back movement and the muscles involved, this would be difficult to

determine and should therefore be approached with caution. This provides an

opportunity for further research, especially since the effects of side reins and other

training aids in the literature is under studied.

7.3 Conclusions

This study has succeeded in its overall aim of making a quantification of the

locomotory biomechanics of the axial skeleton of the horse when trotting on an

aqua-treadmill, using the most state-of-the-art gait analysis technologies. The four

chapters of this thesis answered the four main aims which were to; determine the

impact of water depth on vertical pelvis and wither displacement while trotting on

an aqua-treadmill (Chapter 3), determine the effect of side reins on vertical

displacements of the pelvis and withers when trotting on an aqua-treadmill at

increasing water depths (Chapter 4), determine the impact of water depth on

mediolateral pelvis and wither displacements when trotting on an aqua-treadmill

(Chapter 5), and to determine the difference a few centimetres of water makes to

pelvis and wither displacements when trotting on the aqua-treadmill compared to

overground trotting (Chapter 6). This project has provided useful, quantifiable

information which was previously lacking in scientific literature and non-existent

with regards to aqua-treadmill studies.

215

Water was found to have a significant effect on several locomotory parameters. In

particular, vertical displacements of the pelvis and withers increased as water

depth increased and vertical displacements of the pelvis were significantly greater

than those of the withers. A mean (±SEM) of 115.25 (±2.52) mm for the pelvis,

compared to a mean (±SEM) of 91.53 (±2.36) mm for the withers. Also, the pelvis

exhibited greater minimum and maximum vertical parts of the stride than the

withers. These increases in displacement are indicative of greater compression of

the limbs and likely therefore, indicative of greater workloads at deeper water

depths. Trotting in a small amount of water on an aqua-treadmill induced a mean

(±SEM) increase in vertical displacements of the pelvis 28.18 (±4.51) mm

compared to trotting overground, believed to be due to tactile stimulation of the

water. Side reins had no direct effect on the vertical or mediolateral displacements

during aqua-treadmill exercise (with a maximum water depth of mid carpus) so it

can be deduced that side reins have no direct impact on the work load and

associated changes in parameters measured here, although it is likely that other

muscles are engaged and is therefore an area requiring further investigation. The

aqua-treadmill may be beneficial in increasing the work load for the horse as it is

suggested that changing mechanics are likely to influence loading that could

influence strength, endurance and muscle mass but without detrimental effects to

cranial-caudal or mediolateral symmetry patterns.

Practical experience of training and exercising horses in conjunction with

knowledge of this primary scientific data will enable aqua-treadmill operators to

better assist and benefit the horses they are working with whilst also standardising

216

exercise protocols at the same time as personalising the exercise for the individual

subject.

7.4 Future research and limitations

It is important to recognise the limitations of any study where appropriate and

where suggestions can be made for improvement. Overall, this study has

produced some informative primary data but as always, there are areas for

improvement. The number of final complete sets of data from the trials was a little

disappointing. Initially, for the Qualisys© data collection trial, ten horses were put

forward and began the trials, and for the Xsens© data collection period, fifteen

horses began the trials, which would have given an overall 25 horses of riding

school type all habituated to aqua-treadmill exercise. Unfortunately, due to

various reasons, most often equipment malfunction, there were only a final seven

sets of complete data from the Qualisys© trials and ten from the Xsens© trials.

One horse was withdrawn during the later stages of the exercise protocol due to

fatigue, and with one horse in the second last stage of the exercise protocol on the

aqua-treadmill, there was a power-cut, meaning that the trial could not be

completed. It was deemed important to use only the horses with full and complete

sets of data for accuracy in statistical analyses, with each horse being able to act

as its own control in each trial. Of course, it would have also been ideal to have

been able to select a cohort of horses of all the same breed, type, age and

perhaps more specifically, height, due to the variations seen in vertical

displacements from different height horses. A further interesting study would be to

investigate more fully if height of horse affects the vertical displacement

217

amplitudes they can achieve and therefore the impact of water depth. Although a

more comprehensive study on the conformation of the whole horse may be more

appropriate as the vertical height a horse can reach during a trot stride is more

likely due to its physical make-up, breeding and training.

One paper recognised a specific potential technical issue with the Xsens© system

where it is considered that connecting several sensors in series to a single XBus

Master in a daisy chain style causes a bias error in the magnetometers, which

results in up to a 15° error in heading but only if nine extra IMUs are added as this

creates a local magnetic field, but the ‘daisy chain’ effect can be mitigated by

correctly calibrating the sensors (Brodie et al., 2008). However, the issue of

magnetic field is possibly very apt in this current study, as data were collected

within the confines of the aqua-treadmill environment which is a galvanized steel

metal box. It is possible that the metal from the aqua-treadmill itself had an effect

on the magnetometers in the inertial measurement units, and possibly the

accelerometers and gyroscopes too although to a lesser extent. This may account

for the larger roll amplitudes reported in this study compared to overground

studies. Roll amplitudes reported overground in this study were larger than in

previous literature, but that may be accounted for by the physical transferring of

the IMUs from overground to the aqua-treadmill for each horse, and actually the

IMUs may need re-calibration after each exposure to the stainless-steel

environment of the aqua-treadmill. A validation of Xsens© against Qualisys© was

performed on the vertical displacement data (Chapter 2.5), although there do not

currently appear to be any reports in the literature validating either piece of gait

analysis equipment in the novel environment of an aqua-treadmill. This therefore

218

requires further specific investigation before any fundamental procedural changes

to aqua-treadmill exercise should be made based on the output from this study.

A significant area for improvement in this study would be to make a direct

comparison between treadmill locomotion on a dry belt and aqua-treadmill

locomotion with just a small amount of water. Unfortunately, at the time of data

collection, we were misinformed about the operation of the aqua-treadmill and

were instructed to ensure that there was always at least a few centimetres of water

over the treadmill belt, hence our baseline water depth of mid P3. However, aqua-

treadmills can in fact be run with no apparent water on the belt but as the belt is

submerged in its own water bath, as long as the belt comes through wet, that is

perfectly safe and acceptable operation (Beckmann, 2016, personal

correspondence). A further, more accurate investigation could then be made into

the effects of water on locomotion. It would also be useful to perhaps further

validate treadmills in their own right, as there are now several manufacturers of

treadmills, including high-speed and aqua-treadmills. There needs to be some

standardisation of stated speeds, including calibrations to determine accuracy in

maintaining speed, plus some quantification and calibration of the added effect of

the weight of the water plus the combined weight of the water and the horse on the

speeds of the aqua-treadmill, especially as it has been reported that on a dry high-

speed treadmill, the speed of the belt is reduced as much as 9% during stance

(Buchner et al., 1994a), so it would be expected that the heavier the weight on the

belt the greater the reduction in speed during stance, perhaps dependent upon the

power output of the drive motor. Validation and standardisation of treadmill

equipment would therefore seem a necessity before further accurate locomotory

studies can take place.

219

During the Xsens© trial, horses were trotted in-hand overground first to collect

overground data and to establish an apparent chosen speed of each horse

overground. These speeds were matched on the aqua-treadmill for each horse,

as previous research has established that the optimum energy preserving system

of the spring-mass model is affected by velocity (Blickhan, 1989; Farley et al.,

1993). In the Qualisys© trial it was not possible to match the horse’s speeds on

the aqua-treadmill to their speeds overground as no overground experimental

protocol was included due to time constraints of using the equipment. Ideally, this

would have taken place, however, the two systems were successfully validated

against each other.

Another useful area for investigation, and mentioned previously in this thesis, is

the effectiveness and calibration of hoof or distal limb mounted accelerometers.

Some previous studies in the literature have employed this methodology, but it

would be pertinent to investigate accelerometers in water as there may be some

doubt over the efficacy of the data produced as there are likely different forces

acting on the accelerometers due to the wash of the water. If hoof or distal limb

mounted accelerometers can be validated, then a complementary study with

appendicular mounted IMUs as well would be very useful.

This study only employed an exercise protocol that included a water depth of up to

mid carpus. It was deemed that the riding school type horses that were used in

the study would not be able to manage exercising in deeper water, which is

frustrating as many of the previous studies in the literature feature both walking

and trotting at very deep water depths such as elbow, point of shoulder and stifle.

However, from both personal observation, and now reports in the literature, horses

220

do have a change in locomotion patterns when water becomes so deep that the

horse can no longer lift and step over the water but start to have to wade (Mooij et

al., 2013; Nankervis et al., 2016). However, these recent studies have so far only

investigated walk with the more sophisticated gait analysis technologies leaving

scope for further investigations in this area.

Arguably the largest and most understudied area therefore requiring investigation

is the long-term effects of aqua-treadmill exercise at different water depths.

Longitudinal studies are therefore of imperative importance and should be seen as

a priority. As yet, there are no reports in the literature of contra-indications of

using an aqua-treadmill, which of course, is positive, but further research must

identify specific cases for rehabilitation in an attempt to propose ideal aqua-

treadmill protocols to treat specific conditions. This thesis, therefore, is a mere

starting point for this type of work having quantified the effect of water depth on

certain locomotory parameters and having quantified the potentially beneficial

effects of using side reins. It must now be put forward how to best use the

knowledge of these locomotory parameters in specific instances.

7.5 Implications and applicability to the equine industry

The implications of this study for the equine industry are useful and important.

Quantification of locomotory parameters when exercising through water at the trot

are currently unique and add to existing scientific literature that has investigated

locomotory parameters at the walk. Significantly, it seems that anyone (with

enough funds) can purchase and operate an aqua-treadmill with potentially little or

221

no knowledge of equine locomotion, biomechanics and positive and potential

negative effects of different types of exercise in different situations. It has been

previously reported in the literature that not all experienced Veterinary Clinicians

can agree on lamenesses so it is more likely that a layperson might miss

something small, but potentially significant in a horse’s way of going on an aqua-

treadmill. Thereby, instead of working the aqua-treadmill to alleviate the problem,

there is the potential to make a condition worse. Currently, there are no known

contra-indications reported in the literature of aqua-treadmill exercise but with the

increasing number of aqua-treadmills in the public domain, it is important that all

users are educated not only in the best possible practices, to ensure safety of

handlers, operators and of course, the horses, but to be educated in how water

acts to change a horse’s way of going.

Aqua-treadmills are not often purchased and operated by Veterinary Practices and

most often not by Veterinary allied professionals either. Producing significant

research such as this study, and getting it into peer-reviewed academic literature

is a necessity, so that it can be readily available to Veterinary Surgeons and allied

scientific professionals so that they can relay this useful information on gait,

locomotion and biomechanical features to their clients and colleagues, who can all

make educated judgements on how different water depths will most benefit the

condition they are trying to treat. It is therefore paramount to produce most of the

work of this thesis for publication in peer-reviewed journals now as a matter of

priority. There is also an apparent lack in the industry of an appropriate governing

body for equine hydrotherapy, where for dogs, the Canine Hydrotherapy

Association (CHA, 2018) exists that aids standardisation in procedures and

protocols, so the development of an equine equivalent would be highly relevant.

222

REFERENCES

ACPAT (2017) Association of Chartered Physiotherapists in Animal Therapy. Available from: http://www.acpat.org/ [Accessed July 2017]. Adair, H. S., (2011) Aquatic therapy for conditioning and treatment of tendon and ligament injuries. Proceedings of the 57th Annual Convention of the American Association of Equine Practitioners. San Antonio, Texas, USA, 18-22 November, 181-185. Allen, K., Johns, S., Hyman, S. S., Sislak, M. D., Davis, S. and Amory, J., (2010) How to diagnose and treat back pain in the horse. AAEP Proceedings, 56, 384-388. Anon (2017) FEI Dressage Rules. Available from: https://inside.fei.org/sites/default/files/DRE-Rules_2017_GA_approved_clean.pdf [Accessed July 2017]. Arkell, M., Archer, R. M., Guitian, F. J., May, S. A., (2006) Evidence of bias affecting the interpretation of the results of local anaesthetic nerve blocks when assessing lameness in horses. Veterinary Record, 159, 346–349. Asheim, A., Knudsen, O., Lindholm, A., Rülcker, C., and Saltin, B., (1970) Heart rates and blood lactate concentrations of standardbred horses during training and racing. Journal of the American Veterinary Medical Association, 157(3), 304-12. Audigié, F., Pourcelot, P., Degueurce, C., Denoix, J. M., and Geiger, D., (1999) Kinematics of the equine back: flexion‐extension movements in sound trotting horses. Equine Veterinary Journal, 31(S30), 210-213. Audigie, F., Pourcelot, P., Degueurce, C., Geiger, D. and Denoix, J. M., (2002) Fourier analysis of trunk displacements: a method to identify the lame limb in trotting horses. Journal of Biomechanics, 35(9), 1173-1182. Auer, J. A., (1989) Physical therapy and rehabilitation. In: Jones, W. E., Equine Sports Medicine. Malvern PA: Lea and Febiger, 249-251. Auty, I. and Linington-Payne, M. (2013) BHS complete training manual for stage 2. 2nd edn. Shrewsbury: Kenilworth Press. Back, W. and Clayton, H. M., (2007) Equine Locomotion. W.B. Saunders. Back, W., Schamhardt, H. C., and Barneveld, A., (1997) Kinematic comparison of the leading and trailing fore‐and hindlimbs at the canter. Equine Veterinary Journal, 29(S23), 80-83.

http://www.ncbi.nlm.nih.gov/pubmed?term=Asheim%20A%5BAuthor%5D&cauthor=true&cauthor_uid=5448842

http://www.ncbi.nlm.nih.gov/pubmed?term=Knudsen%20O%5BAuthor%5D&cauthor=true&cauthor_uid=5448842

http://www.ncbi.nlm.nih.gov/pubmed?term=Lindholm%20A%5BAuthor%5D&cauthor=true&cauthor_uid=5448842

http://www.ncbi.nlm.nih.gov/pubmed?term=R%C3%BClcker%20C%5BAuthor%5D&cauthor=true&cauthor_uid=5448842

http://www.ncbi.nlm.nih.gov/pubmed?term=Saltin%20B%5BAuthor%5D&cauthor=true&cauthor_uid=5448842

223

Back, W., Schamhardt, H. C., Hartman, W. and Barneveld, A. (1995a) Kinematic differences between the distal portions of the forelimbs and hind limbs of horses at the trot. American Journal of Veterinary Research, 56(11), 1522-1528. Back, W., Schamhardt, H. C., Savelberg, H. H. C. M., Bogert, A. J., Bruin, G., Hartman, W. and Barneveld, A., (1995b) How the horse moves: 2. Significance of graphical representations of equine hind limb kinematics. Equine Veterinary Journal, 27(1), 39-45. Barr, A. R., Dow, S. M., and Goodship, A. E., (1995) Parameters of forelimb ground reaction force in 48 normal ponies. The Veterinary Record, 136(12), 283-286. Barrett, E. L., Talbot, A. M., Driver, A. J., Barr, F. J. and Barr, A. R. S., (2006) A technique for pelvic radiography in the standing horse. Equine Veterinary Journal, 38(3), 266-270. Barrey, E., (1990) Investigation of the vertical hoof force distribution in the equine forelimb with an instrumented horseboot. Equine Veterinary Journal, 22(S9), 35-38. Barrey, E., Galloux, P., Valette, J. P., Auvinet, B., and Wolter, R., (1993) Stride characteristics of overground versus treadmill locomotion in the Saddle Horse. Acta Anatomica, 146(2-3), 90-94. Baumann, M., (2015) Personal correspondence on the application of the aqua-treadmill with regards to specific injuries with Dr Matthias Baumann. Advocate of aqua-treadmill exercise in associate with ActivoMed. Beckmann, W., (2016) Personal correspondence on the engineering of the FMBs ActivoMed Aqua-treadmill with Wulf Beckmann, CEO of Activomed, December 2016. BEVA (2017) Musculo-Skeletal Paraprofessionals. British Equine Veterinary Association. Available at: https://www.beva.org.uk/Home/Working-For-Change/Allied-Professionals. [Accessed July 2017]. BHS, (2017) BHS Stage 2 Syllabus. Available from the BHS website http://www.bhs.org.uk/education/exams-and-qualifications/find-exams/exams/stage-2 [Accessed July 2017] Biau, S., Couve, O., Lemaire, S. and Barrey, E., (2002) The effect of reins on kinetic variables of locomotion. Equine Veterinary Journal, 34(S34), 359-362. Biewener, A. A., (1983) Allometry of quadrupedal locomotion: the scaling of duty factor, bone curvature and limb orientation to body size. Journal of Experimental Biology, 105(1), 147-171.

http://www.bhs.org.uk/education/exams-and-qualifications/find-exams/exams/stage-2

http://www.bhs.org.uk/education/exams-and-qualifications/find-exams/exams/stage-2

224

Biewener, A. A., (1998). Muscle-tendon stresses and elastic energy storage during locomotion in the horse. Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology, 120(1), 73-87. Bleakley, C. M., and Davison, G. W., (2009) What is the biochemical and physiological rationale for using cold-water immersion in sports recovery? A systematic review. British Journal of Sports Medicine, 44, 179-187. Blickhan, R., (1989) The spring-mass model for running and hopping. Journal of Biomechanics, 22(11-12), 1217-1227. Bonin, S.J., Clayton, H.M., Lanovaz, J.L. and, Johnston, T., (2007) Comparison of mandibular motion in horses chewing hay and pellets. Equine Veterinary Journal, 39(3), 258-262. Borgia, L. A., Valberg, S. J., and Essen-Gustavsson, B., (2010) Differences in the metabolic properties of gluteus medius and superficial digital flexor muscles and the effect of water treadmill training in the horse. Equine Veterinary Journal, 42(S38), 665-670. Brodie, M. A., Walmsley, A. and Page, W., (2008) The static accuracy and calibration of inertial measurement units for 3D orientation. Computer Methods in Biomechanics and Biomedical Engineering, 11(6), 641-648. Bromiley, M. W., (2000) Physical Therapy in Equine Veterinary Medicine: Useful or Useless? Proceedings of the Annual Convention of the AAEP, 46, 94-97. Buchner, H. H. F., Kastner, J., Girtler, D., and, Knezevic, P. F., (1993) Quantification of Hind Limb Lameness in the Horse. Cells Tissues Organs, 146, 196-199. Buchner, H. H. F., Obermüller, S. and Scheidl, M., (2000) Body centre of mass movement in the sound horse. The Veterinary Journal, 160(3), 225-234. Buchner, H. H. F., and Schildboeck, U., (2006) Physiotherapy Applied to the Horse: A Review. Equine Veterinary Journal, 38(6), 574-580. Buchner, H. H. F., Savelberg, H. H. C. M., Schamhardt, H. C., Merkens, H. W., and Barneveld, A., (1994a) Kinematics of Treadmill Versus Overground Locomotion in Horses. The Veterinary Quarterly, 16(S2), 84-90. Buchner, H. H. F., Savelberg, H. H. C. M., Schamhardt, H. C., Merkens, H. W., and Barneveld, A., (1994b) Habituation of horses to treadmill locomotion. Equine Veterinary Journal, 26(S17), 13-15.

225

Buchner, H. H. F., Savelberg, H. H. C. M., Schamhardt, H. C. and Barneveld, A., (1995) Temporal stride patterns in horses with experimentally induced fore‐or hindlimb lameness. Equine Veterinary Journal, 27(S18), 161-165. Buchner, H. H. F., Savelberg, H. H. C. M., Schamhardt, H.C. and Barneveld, A., (1996a) Head and trunk movement adaptations in horses with experimentally induced fore‐or hindlimb lameness. Equine Veterinary Journal, 28(1), 71-76. Buchner, H. H. F., Savelberg, H. H. C. M., Schamhardt, H.C. and Barneveld, A., (1996b) Limb movement adaptations in horses with experimentally induced fore‐or hindlimb lameness. Equine Veterinary Journal, 28(1), 63-70. Burn, J. F. and, Brockington, C., (2001) Quantification of hoof deformation using optical motion capture. Equine Veterinary Journal, 33(S33), 50-53. Burns, G., Dart, A., Jeffcott, L., (2016) Clinical progress in the diagnosis of thoracolumbar problems in horses. Equine Veterinary Education. DOI: 10.1111/eve.12623 Early View (Online Version of Record published before inclusion in an issue). Butler, J. A., Colles, C. M., Dyson, S. J., Kold, S. E. and Poulos, P. W., (2000) The Spine, in Clinical Radiology of the Horse, Second Edition, Blackwell Science Ltd, Oxford, UK. Cadiot, P. J. and Almy, J., (1924) Affections des régions et des organes suite. In: Traité de Thérapeutique Chirurgicale des animaux domestiques. Asselin et Houzeau, Paris. Camus, M., Chateau, H., Holden-Douilly, L., Robin, D., Falala, S., Ravary-Plumiöen, B., Pourcelot, P. and Crevier-Denoix, N., (2012) Use of a 3D dynamometric horseshoe for the measurement of grip parameters in a horse cantering on right and left circles on two surfaces. Computer Methods in Biomechanics and Biomedical Engineering, 15(S1), 132-134. CET Equine Spas (2017) CET Equine Spa. Available from: http://www.cet-equine-spa.com/ [Accessed July 2017] CHA, (2018) Canine Hydrotherapy Association. Available from: http://www.canine-hydrotherapy.org/ [Accessed May 2018]. Chateau, H., Robin, D., Simonelli, T., Pacquet, L., Pourcelot, P., Falala, S., Denoix, J.M. and Crevier-Denoix, N., (2009) Design and validation of a dynamometric horseshoe for the measurement of three-dimensional ground reaction force on a moving horse. Journal of Biomechanics, 42(3), 336-340.

226

Chung, W. M., Yeung, S., Chan, W. W. and Lee, R., (2011) Validity of VICON motion analysis system for upper limb kinematic measurement–A comparison study with inertial tracking Xsens system. Hong Kong Physiotherapy Journal, 29(2), 97. Church, E. E., Walker, A. M., Wilson, A. M. and Pfau, T., (2009) Evaluation of discriminant analysis based on dorsoventral symmetry indices to quantify hindlimb lameness during over ground locomotion in the horse. Equine Veterinary Journal, 41(3), 304-308. Clayton, H. M., (1989) Locomotion. In: Jones, W.E. Equine Sports Medicine. USA: Lea & Febiger pp.149-187. Clayton, H. M., (1994a), Comparison of the stride kinematics of the collected, working, medium and extended trot in horses. Equine Veterinary Journal, 26, 230–234. Clayton, H. M., (1994b) Comparison of the stride kinematics of the collected, working, medium and extended canter in horses. Equine Veterinary Journal, 26(S17), 16-19. Clayton, H. M., (1995) Comparison of the stride kinematics of the collected, medium, and extended walks in horses. American Journal of Veterinary Research, 56(7), 849-852. Clayton, H. M., Larson, B., Kaiser, L. J. and Lavagnino, M., (2011) Length and elasticity of side reins affect rein tension at trot. The Veterinary Journal, 188(3), 291-294. Clayton, H.M., Lavagnino, M., Kaiser, L.J. and Stubbs, N.C., (2011) Evaluation of biomechanical effects of four stimulation devices placed on the hind feet of trotting horses. American Journal of Veterinary Research, 72(11), 1489-1495. Clayton and Schamhardt, (2007) In: Back, W. and Clayton, H. M. (2007) Equine Locomotion. Second Edition. W.B. Saunders. Clayton, H. M., Singleton, W. H., Lanovaz, J. L. and Cloud, G. L., (2003) Measurement of rein tension during horseback riding using strain gage transducers. Experimental Techniques, 27(3), 34-36. Clayton, H. M., Singleton, W. H., Lanovaz, J. L. and Cloud, G. L., (2005) Strain gauge measurement of rein tension during riding: a pilot study. Equine and Comparative Exercise Physiology, 2(03), 203-205. Clayton, H. M., and Townsend, H. G. G., (1989) Kinematics of the cervical spine of the adult horse. Equine Veterinary Journal, 21(3), 189-192.

227

Clayton, H. M., White, A. D., Kaiser, L. J., Nauwelaerts, S., Lavagnino, M. and Stubbs, N. C., (2008) Short-term habituation of equine limb kinematics to tactile stimulation of the coronet. Veterinary and Comparative Orthopaedics and Traumatology, 21(3), 211. Clayton, H. M., White, A. D., Kaiser, L. J., Nauwelaerts, S., Lavagnino, M. and Stubbs, N. C., (2010) Hindlimb response to tactile stimulation of the pastern and coronet. Equine Veterinary Journal, 42(3), 227-233. Corley, J. M. and Goodship, A. E., (1994) Treadmill training induced changes to some kinematic variables measured at the canter in Thoroughbred fillies. Equine Veterinary Journal, 26(S17), 20-24. Cottriall, S., Ritruechai, P. and Wakeling, J. M., (2009) The effects of training aids on the longissimus dorsi in the equine back. Comparative Exercise Physiology, 5(3-4), 111-114. Crevier-Denoix, N., Camus, M., Falala, S., Ravary-Plumioen, B., Pauchard, M., Martino, J., Desquilbet, L., Chateau, H. and Pourcelot, P., (2014) 3D measurement of the displacement of the forelimb hoof during stance in three horses circling at the canter on a competition surface. Computer Methods in Biomechanics and Biomedical Engineering, 17(S1), 142-143. Cureton, K. J., (1997) Physiologic responses to water exercise. In: Routi, R. G., and Morris, D. M., (eds) Aquatic Rehabilitation. Lippincott: Philadelphia, 39-46. Cutti, A.G., Giovanardi, A., Rocchi, L., Davalli, A. and Sacchetti, R., (2008) Ambulatory measurement of shoulder and elbow kinematics through inertial and magnetic sensors. Medical & Biological Engineering & Computing, 46(2), 169-178. Denning, W. M., Bressel, E., Dolny, D., Bressel, M., and Seeley, M. K., (2012) A Review of Biophysical Differences Between Aquatic and Land-Based Exercise. International Journal of Aquatic Research and Education, 6, 46-67. Denoix, J. M., (1999) Spinal biomechanics and functional anatomy. The Veterinary Clinics of North America Equine Practice, 15(1), 27-60. Denoix, J. M. and Audigie, F., (2001) The neck and back. In: Equine Locomotion, First Edition, Eds: W. Back and H. Clayton, W.B. Saunders, London. pp 167-191. Denoix, J., Pailloux, J. and Lewis, J. (tr.) (2001) Physical therapy and massage for the horse. Second Edition. London: Manson Publishing. Deuel, N.R. and Lawrence, L.M., (1987) Laterality in the gallop gait of horses. Journal of Biomechanics, 20(6), 645-649.

228

Deuel, N. R. and Park, J., (1990) Canter lead change kinematics of superior Olympic dressage horses. Journal of Equine Veterinary Science, 10(4), 287-298. Dimery, N. J., Alexander, R. M. and Ker, R. F., (1986) Elastic extension of leg tendons in the locomotion of horses (Equus caballus). Journal of Zoology, 210(3), 415-425. Drevemo, S., Fredricson, I., Hjerten, G. and McMiken, D., (1987) Early development of gait asymmetries in trotting Standardbred colts. Equine Veterinary Journal, 19(3), 189-191. Dutto, D. J., Hoyt, D. F., Clayton, H. M., Cogger, E. A., and Wickler, S. J., (2006) Joint work and power for both the forelimb and hindlimb during trotting in the horse. Journal of Experimental Biology, 209(20), 3990-3999. Dyson, S., (2011), Can lameness be graded reliably? Equine Veterinary Journal, 43, 379–382. Equestrian and Horse (2017a) Training Aid: Chambon Available from: http://www.equestrianandhorse.com/training-equipment/chambon.html [Accessed July 2017] Equestrian and Horse (2017b) Training Aid: de Gogue Available from: http://www.equestrianandhorse.com/training-equipment/de-gogue.html [Accessed July 2017] Erichsen, C., Eksell, P., Holm, K. R., Lord, P., and Johnston, C., (2004) Relationship between scintigraphic and radiographic evaluations of spinous processes in the thoracolumbar spine in riding horses without clinical signs of back problems. Equine Veterinary Journal, 36(6), 458-465. Erichsen, C., (2003) Diagnostic imaging of the equine thoracolumbar spine and sacroiliac joint region. Norwegian College of Veterinary Medicine, Department of Large Animal Clinical Sciences. Submitted as a Thesis for the degree of Doctor Medicinae Veterinariae. Available from: http://equirad.no/wp-content/uploads/2011/04/CEE_Avhandling.pdf [Accessed July 2017]. Faber, M., Johnston, C., Weeren, P. V., and Barneveld, A., (2002) Repeatability of back kinematics in horses during treadmill locomotion. Equine Veterinary Journal, 34(3), 235-241. Faber, M., Johnston, C., Schamhardt, H., Weeren, R. V., Roepstorff, L. and Barneveld, A., (2001a) Basic three-dimensional kinematics of the vertebral column of horses trotting on a treadmill. American Journal of Veterinary Research, 62(5), 757-764.

229

Faber, M., Johnston, C., Schamhardt, H.C., Weeren, P.V., Roepstorff, L. and Barneveld, A., (2001b) Three‐dimensional kinematics of the equine spine during canter. Equine Veterinary Journal, 33(S33), 145-149. Faber, M., Schamhardt, H., Weeren, R. V. and Barneveld, A., (2001c) Methodology and validity of assessing kinematics of the thoracolumbar vertebral column in horses on the basis of skin-fixated markers. American Journal of Veterinary Research, 62(3), 301-306. Faber, M., Schamhardt, H., Weeren, R. V., Johnston, C., Roepstorff, L. and Barneveld, A. B., (2000) Basic three-dimensional kinematics of the vertebral column of horses walking on a treadmill. American Journal of Veterinary Research, 61(4), 399-406. Farley, C. T., Glasheen, J. and McMahon, T. A., (1993) Running springs: speed and animal size. Journal of Experimental Biology, 185(1), 71-86. Findley, J. and Singer, E., (2015) Equine back disorders 1. Clinical presentation, investigation and diagnosis. In Practice, 37(9), 456-467. Findlow, A., Goulermas, J. Y., Nester, C., Howard, D., and Kenney, L.P.J., (2008) Predicting lower limb joint kinematics using wearable motion sensors. Gait & Posture, 28(1), 120-126. Firshman, A. M., Borgia, L. A. and Valberg, S. J., (2015) Effects of training at a walk on conventional and underwater treadmills on fiber properties and metabolic responses of superficial digital flexor and gluteal muscles to high-speed exercise in horses. American Journal of Veterinary Research, 76(12), 1058-1065. FMBs (2017) FMBs Therapy Systems. Available from: https://www.fmbs.co.uk/ [Accessed July 2017]. Frederick, Jr, F. H., and Henderson, J. M., (1970) Impact force measurement using preloaded transducers. American Journal of Veterinary Research, 31, 2279-2283. Fredricson, I., Drevemo, S., Dalin, G., Hjerten, G. and Björne, K., (1980) The application of high‐speed cinematography for the quantitative analysis of equine locomotion. Equine Veterinary Journal, 12(2), 54-59. Fredricson, S, Drevemo, S., Dalin, G., Hjertén, G., Björne, K., Rynde, R., and Franzen, G., (1983) Treadmill for Equine Locomotion Analysis. Equine Veterinary Journal, 15(2), 111-115.

230

Fuller, C.J., Bladon, B.M., Driver, A.J. and Barr, A.R., (2006) The intra-and inter-assessor reliability of measurement of functional outcome by lameness scoring in horses. The Veterinary Journal, 171(2), 281-286. Galisteo, A. M., Cano, M. R., (Morales, J. L., Vivo, J., and Miro, F., (1998) The influence of speed and height at the withers on the kinematics of sound horses at the handled trot. Veterinary Research Communications, 22(6), 415-423. GFS, (2017) Fieldhouse Riding Equipment (2016) Pessoa Training Aid. Available from: http://www.gfsriding.co.uk/products/product.php?pid=10119 [Accessed July 2017] Gomez Álvarez, C. B., Bobbert, M. F., Lamers, L., Johnston, C., Back, W. and Van Weeren, P. R., (2008). The effect of induced hindlimb lameness on thoracolumbar kinematics during treadmill locomotion. Equine Veterinary Journal, 40(2), 147-152. Gomez Álvarez, C. B., Rhodin, M., Bobbert, M. F., Meyer, H., Weishaupt, M. A., Johnston, C. and Van Weeren, P. R., (2006) The effect of head and neck position on the thoracolumbar kinematics in the unridden horse. Equine Veterinary Journal, 38(S36), 445-451. Gomez Álvarez, C. B., Wennerstrand, J., Bobbert, M. F., Lamers, L., Johnston, C., Back, W. and Van Weeren, P. R., (2007) The effect of induced forelimb lameness on thoracolumbar kinematics during treadmill locomotion. Equine Veterinary Journal, 39(3), 197-201. Goody, P.C. and Goody, J., (2000) Horse anatomy: a pictorial approach to equine structure. Second edition. London: J.A. Allen. Greve, L., Pfau, T. and Dyson, S., (2017) Thoracolumbar movement in sound horses trotting in straight lines in hand and on the lunge and the relationship with hind limb symmetry or asymmetry. The Veterinary Journal, 220, 95-104. Greve, L. and Dyson, S., (2014) The interrelationship of lameness, saddle slip and back shape in the general sports horse population. Equine Veterinary Journal, 46, 687-694. Hammarberg M, Egenvall A, Pfau T, Rhodin M., (2016) Rater agreement of visual lameness assessment in horses during lungeing. Equine Veterinary Journal, 48(1), 78–82. Haussler, K. K., Bertram, J. E. A., Gellman, K. and Hermanson, J. W., (2001) Segmental in vivo vertebral kinematics at the walk, trot and canter: a preliminary study. Equine Veterinary Journal, 33(S33), 160-164.

231

Hawksmoor (2017) Level 3 Certificate in Hydrotherapy for Small Mammals. Available from: http://www.hawksmoorhydrotherapy.com/hawksmoor-hydrotherapy-training-courses/accredited-abc-courses/level-3-certificate-in-hydrotherapy-for-small-animals/ [Accessed July 2017] Heleski, C. R., McGreevy, P. D., Kaiser, L. J., Lavagnino, M., Tans, E., Bello, N. and Clayton, H. M., (2009) Effects on behaviour and rein tension on horses ridden with or without martingales and rein inserts. The Veterinary Journal, 181(1), 56-62. Hewetson, M., Christley, R.M., Hunt, I.D. and Voute, L.C., (2006) Investigations of the reliability of observational gait analysis for the assessment of lameness in horses. The Veterinary Record, 158(25), 852-857. Hildebrand, M. (1965) Symmetrical gaits of horses Science, 150(697), 701-8. Hildebrand, M., (1977) Analysis of asymmetrical gaits. Journal of Mammalogy, 58(2), 131-156. Hinman, R. S., Heywood, S. E., and Day, A. R., (2007) Aquatic physical therapy for Hip and Knee Osteoarthritis: Results of a Single-Blind Randomized Controlled Trial. Journal of the American Physical Therapy Association, 87(1), 32-43. Hobo, S., Yoshida, K., and Yoshihara, T., (1998) Characteristics of Respiratory Function during Swimming Exercise in Thoroughbreds. The Journal of Veterinary Medical Science, 60(6), 687-689. Hodgson, D.R., McKeever, K.H. and McGowan, C.M. (2014) The athletic horse: principles and practice of equine sports medicine. 2nd edn. St. Louis: Saunders Elsevier. Hoyt, D. F., and Taylor, C. R., (1981) Gait and the energetics of locomotion in horses. Nature, 292(5820), 239-240. Hoyt, D. F., Wickler, S. J., Dutto, D. J., Catterfeld, G. E. and Johnsen, D., (2006) What are the relations between mechanics, gait parameters, and energetics in terrestrial locomotion? Journal of Experimental Zoology Part A: Ecological Genetics and Physiology, 305(11), 912-922. Hunt, E. R., (2001) Response of twenty-seven horses with lower leg injuries to cold spa bath hydrotherapy Journal of Equine Veterinary Science, 21(4), 188-193. IBM Corp. Released 2013. IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY: IBM Corp. [used 2015-2017].

232

Irwin, D. H. G., and Howell, D. W., (1980) Some thoughts on swimming horses in a pool. Journal of the South African Veterinary Association, 51(3), 189-191. Jackson, R., (1990) Waters and Spas in the Classical World. Medical History, 34(S10), 1-13. Jeffcott, L. B., (1979a) Radiographic Examination of the Equine Vertebral Column. Veterinary Radiology, 20(3-6), 135-139. Jeffcott, L. B., (1979b) Radiographic features of the normal equine thoracolumbar spine. Veterinary Radiology & Ultrasound, 20(3‐6), 140-147. Jeffcott, L.B., (1979c). Back problems in the horse—a look at past, present and future progress. Equine Veterinary Journal, 11(3), 129-136. Jeffcott, L. B., (1980) Disorders of the thoracolumbar spine of the horse—a survey of 443 cases. Equine Veterinary Journal, 12(4), 197-210. Johnston, C., Holm, K. R., Erichsen, C., Eksell, P., and Drevemo, S., (2004) Kinematic evaluation of the back in fully functioning riding horses. Equine Veterinary Journal, 36(6), 495-498. Johnston, C., Holm, K., Faber, M., Erichsen, C., Eksell, P. and Drevemo, S., (2002) Effect of conformational aspects on the movement of the equine back. Equine Veterinary Journal, 34(S34), 314-318. Joly, F., Galoppin, L., Bordat, P., Cousse, H., and Neuzil, E., (2000) Calcium and bicarbonate ions mediate the inhibition of mast cell histamine release by Avène spa water. Fundamental and Clinical Pharmacology, 14(6), 611-613. Jones, W.E., (1988) Equine sports medicine. USA: Lea & Febiger. Jones, J. H., and Hiraga, A., (2006) Metabolic, Cardiovascular and Respiratory Responses to Swimming in Horses. Proceedings of a Workshop on Exercise Induced Pulmonary Haemorrhage: State of current knowledge. 9th – 12th March Granville Island, Vancouver, Canada. Havemeyer Foundation Monograph, Series No. 20. Kai, M., Aoki, O., Hiraga, A., Oki, H. and Tokuriki, M., (2000) Use of an instrument sandwiched between the hoof and shoe to measure vertical ground reaction forces and three-dimensional acceleration at the walk, trot, and canter in horses. American Journal of Veterinary Research, 61(8), 979-985. Kato, T., Ohmura, H., Hiraga, A., Wada, S., Kuwahara, M., and Tsubone, H., (2003) Changes in heart rate variability in horses during immersion in warm springwater. American Journal of Veterinary Research, 64(12), 1482-1485.

233

Keegan, K. G., (2007) Evidence-based lameness detection and quantification. Veterinary Clinics of North America: Equine Practice, 23(2), 403-423. Keegan, K. G., Dent, E. V., Wilson, D. A., Janicek, J., Kramer, J., Lacarrubba, A., Walsh, D. M., Cassells, M. W., Esther, T. M., Schiltz, P. and Frees, K. E., (2010) Repeatability of subjective evaluation of lameness in horses. Equine Veterinary Journal, 42(2), 92-97. Keegan, K. G., Kramer, J., Yonezawa, Y., Maki, H., Pai, P. F., Dent, E. V., Kellerman, T. E., Wilson, D. A. and Reed, S. K., (2011) Assessment of repeatability of a wireless, inertial sensor–based lameness evaluation system for horses. American Journal of Veterinary Research, 72(9), 1156-1163. Keegan, K. G., Pai, P. F., Wilson, D. A. and Smith, B. K., (2001) Signal decomposition method of evaluating head movement to measure induced forelimb lameness in horses trotting on a treadmill. Equine Veterinary Journal, 33(5), 446-451. Keegan, K. G., Wilson, D. A., Wilson, D. J., Smith, B., Gaughan, E. M., Pleasant, R. S., Lillich, J. D., Kramer, J., Howard, R. D., Bacon-Miller, C. and Davis, E. G., (1998) Evaluation of mild lameness in horses trotting on a treadmill by clinicians and interns or residents and correlation of their assessments with kinematic gait analysis. American Journal of Veterinary Research, 59(11), 1370-1377. Keegan, K. G., Wilson, D. A., Kramer, J., Reed, S. K., Yonezawa, Y., Maki, H., Pai, P. F. and Lopes, M. A., (2013) Comparison of a body-mounted inertial sensor system–based method with subjective evaluation for detection of lameness in horses. American Journal of Veterinary Research, 74(1), 17-24. Keegan, K. G., Yonezawa, Y., Pai, P. F., Wilson, D. A. and Kramer, J., (2004) Evaluation of a sensor-based system of motion analysis for detection and quantification of forelimb and hind limb lameness in horses. American Journal of Veterinary Research, 65(5), 665-670. Kelmer, G., Keegan, K. G., Kramer, J., Wilson, D. A., Pai, F. P. and Singh, P., (2005) Computer-assisted kinematic evaluation of induced compensatory movements resembling lameness in horses trotting on a treadmill. American Journal of Veterinary Research, 66(4), 646-655. King, M. R., Haussler, K. K., Kawcak, C. E., McIlwraith, C. W., and Reiser, R. F., (2013a) Mechanisms of aquatic therapy and its potential use in managing equine osteoarthritis. Equine Veterinary Education, 25(4), 204-209. King, M. R., Haussler, K. K., Kawcak, C. E., McIlwraith, C. W., and Reiser, R. F., (2013b) Effect of underwater treadmill exercise on postural sway in horses with experimentally induced carpal joint osteoarthritis. American Journal of Veterinary Research, 74(7), 971-982.

234

Laerd Statistics (2015). Statistical tutorials and software guides. Available from https://statistics.laerd.com/ [Accessed 2016-2017] Leach, D. H., and Drevemo, S., (1991) Velocity-dependent Changes in a Stride Frequency and Length of trotters on a Treadmill. Equine Exercise Physiology, 3, 136-140. Leach, D.H., and Dyson, S., (1988) Instant centres of rotation of equine limb joints and their relationship to standard skin marker locations. Equine Veterinary Journal, 20(S6), 113-119. Lefrancois, K., and Nankervis, K. J., (2016) Distal Limb Movement in Walk on a Treadmill with and Without Water. Equine Veterinary Journal, 48(S49): 24–25. Licka, T. F., Peham, C. and Frey, A., (2004) Electromyographic activity of the longissimus dorsi muscles in horses during trotting on a treadmill. American Journal of Veterinary Research, 65(2), 155-158. Lindner, A., Wäschle, S., and Sasse, H. H. L., (2011) Physiological and blood biochemical variables in horses exercising on a treadmill submerged in water. Journal of Animal Physiology and Animal Nutrition, 10, 1-7. Lindner, A., Wurm, S., Beuttler, J. and Sasse, H.H.L. (2003) Effect of water height on biochemistry and heart rate of horses exercising on a treadmill submerged in water. In: Proceedings of the 18th Equine Nutrition and Physiology Society, June 4-7, Lansing, East Lansing, Michigan. pp 204-206. Linford, R. L., (1994) Camera speeds for optoelectronic assessment of stride-timing characteristics in horses at the trot. American Journal of Veterinary Research, 55(9), 1189-1195. Maliye, S., Voute, L., Lund, D. and Marshall, J. F., (2013) An inertial sensor‐based system can objectively assess diagnostic anaesthesia of the equine foot. Equine Veterinary Journal, 45(S45), 26-30. Marshall, J. F., Lund, D. G. and Voute, L. C., (2012) Use of a wireless, inertial sensor‐based system to objectively evaluate flexion tests in the horse. Equine Veterinary Journal, 44(S43), 8-11. Martin, P., Chateau, H., Pourcelot, P., Duray, L. and Cheze, L., (2014) Comparison Between Inertial Sensors and Motion Capture System to Quantify Flexion‐Extension Motion in the Back of a Horse. Equine Veterinary Journal, 46(S46), 43-43. MATLAB (R14, The Mathsworks INC, Natick, MA, US) Matz, H., Orion, E., and Wolf, R., (2003) Balneotherapy in dermatology. Dermatologic Therapy, 16, 132-140.

235

Mauchly, J. W., (1940) Significance Test for Sphericity of a Normal n-Variate Distribution. The Annals of Mathematical Statistics, 11(2), 204–209. May, S. A. and Wyn-Jones, G., (1987) Identification of hindleg lameness. Equine Veterinary Journal, 19(3), 185-188. McCracken, M. J., Kramer, J., Keegan, K. G., Lopes, M., Wilson, D. A., Reed, S. K., LaCarrubba, A. and Rasch, M., (2012) Comparison of an inertial sensor system of lameness quantification with subjective lameness evaluation. Equine Veterinary Journal, 44(6), 652-656. McGowan, C., (2008) Introduction. In: McGowan, C., Goff, L., Stubbs, N., (2008), Animal Physiotherapy – Assessment, Treatment and Rehabilitation of Animals. Oxford: Blackwell Publishing. Ch. 1. Mendez-Angulo, J. L., Firshman, A. M., Groschen, D. M., Kieffer, P. J., Trumble, T. N., (2013) Effect of water depth on amount of flexion and extension of joints of the distal aspects of the limbs in healthy horses walking on an underwater treadmill. American Journal of Veterinary Research, 74(4), 557-566. Michlovitz, S. L., (1990) Thermal agents in rehabilitation. Philadelphia: FA Davis 63-67. Cited in Hunt, E. R., (2001) Response of twenty-seven horses with lower leg injuries to cold spa bath hydrotherapy Journal of Equine Veterinary Science, 21(4), 188-193. Minetti, A.E., (1998). The biomechanics of skipping gaits: a third locomotion paradigm? Proceedings of the Royal Society of London B: Biological Sciences, 265(1402), 1227-1233. Misumi, K., Hirakawa, A., Sakamoto, H., and Shimizu, R., (1994a) Principal component analysis, using the measurements during running and swimming test, in thoroughbred horses. Journal of Veterinary Medical Science, 56(6),1075-80. Misumi, K., Sakamoto, H., and Shimizu, R., (1994b) Changes in blood lactate and heart rate in thoroughbred horses during swimming and running according to their stage of training. Veterinary Record, 135, 226-228. Misumi, K., Sakamoto, H., and Shimizu, R., (1994c) The validity of swimming training for two-year-old thoroughbreds. The Journal of Veterinary Medical Science, 56(2), 217-222. Misumi, K., Sakamoto, H., and Shimizu, R., (1995) Changes in skeletal muscle composition in response to swimming training for young horses. The Journal of Veterinary Medical Science, 57(5), 959-961. Mooij, M. J. W., Jans, W., den Heijer, de Pater, M., and Back, W., (2013) Biomechanical responses of the back of riding horses to water treadmill exercise. The Veterinary Journal, 198, 120-123.

http://www.ncbi.nlm.nih.gov/pubmed?term=Misumi%20K%5BAuthor%5D&cauthor=true&cauthor_uid=7696395

http://www.ncbi.nlm.nih.gov/pubmed?term=Hirakawa%20A%5BAuthor%5D&cauthor=true&cauthor_uid=7696395

http://www.ncbi.nlm.nih.gov/pubmed?term=Sakamoto%20H%5BAuthor%5D&cauthor=true&cauthor_uid=7696395

http://www.ncbi.nlm.nih.gov/pubmed?term=Shimizu%20R%5BAuthor%5D&cauthor=true&cauthor_uid=7696395

http://www.ncbi.nlm.nih.gov/pubmed?term=principal%20component%20analysis%20swimming%20horses

236

Moulton College, (2014) BTEC Level 4 Professional Award in Canine Hydrotherapy. Details available from http://www.moulton.ac.uk/business/enterprise-at-moulton/animal-therapy-centre/new-document [Accessed October, 2014]. Mourcou, Q., Fleury, A., Franco, C., Klopcic, F. and Vuillerme, N., (2015) Performance evaluation of smartphone inertial sensors measurement for range of motion. Sensors, 15(9), 23168-23187 Muybridge, E., and Mozley, A. V., (1979) Muybridge's complete Human and Animal Locomotion. All 781 Plates from the 1887 ‘Animal Locomotion', 3 Volumes. Dover Publications, Inc., New York. Nankervis, K. J., Finney, P. and Launder, L., (2016) Water depth modifies back kinematics of horses during water treadmill exercise. Equine Veterinary Journal, 48(6), 732–736. Nankervis, K. J., Launder, E. J. and Murray, R. C., (2017) The use of treadmills within the rehabilitation of horses. Journal of Equine Veterinary Science, 53, 108-115. Nankervis, K. J., Thomas, S., and Marlin, D. J., (2008a) Effect of water temperature on heart rate of horses during water treadmill exercise. Comparative Exercise Physiology, 5(3-4), 127-131. Nankervis, K., Hodgins, D. and Marlin, D., (2008b) Comparison between a sensor (3D accelerometer) and ProReflex motion capture systems to measure stride frequency of horses on a treadmill. Comparative Exercise Physiology, 5(3-4), 107-109. Nankervis, K. J., and Williams, R. J., (2006) Heart rate responses during acclimation of horses to water treadmill exercise. Equine Veterinary Journal, 38(S36), 110-112. Nasermoaddeli, A., and Kagamimori, S., (2005) Balneotherapy in Medicine: A Review. Environmental Health and Preventive Medicine, 10(4), 171-179. O’Hare, J. P., Heywood, A., Summerhayes, C., Lunn, G., Evans, J. M., Walters, G., Corrall, R. J. M., and Dieppe, P. A., (1985) Observations on the effects of immersion in Bath spa water. British Medical Journal, 291(6511), 21-28. Old Mill Saddlery (2017) Elastic Bungie Training Rein Available from: http://www.saddlery.biz/elastic-bungie-training-rein?sqr=bungie& [Accessed July 2017].

237

Olsen, E., Haubro Andersen, P., and Pfau, T., (2012) Accuracy and precision of equine gait event detection during walking with limb and trunk mounted inertial sensors. Sensors, 12(6), 8145-8156. Parkes, R. S., Weller, R., Groth, A. M., May, S. and Pfau, T., (2009) Evidence of the development of 'domain-restricted' expertise in the recognition of asymmetric motion characteristics of hindlimb lameness in the horse. Equine Veterinary Journal, 41(2), 112-117. Parsons, K.J. and Wilson, A.M., (2006) The use of MP3 recorders to log data from equine hoof mounted accelerometers. Equine Veterinary Journal, 38(7), 675-680. Parsons, K. J., Pfau, T., Ferrari, M. and Wilson, A. M., (2008) High-speed gallop locomotion in the Thoroughbred racehorse. II. The effect of incline on centre of mass movement and mechanical energy fluctuation. Journal of Experimental Biology, 211(6), 945-956. Pauchard, M., Chateau, H., Camus, M., Ravary-Plumioen, B., Falala, S., Martino, J., Desquilbet, L., Denoix, J. M., Pourcelot, P., and Crevier-Denoix, N., (2014) Effects of speed on the vertical amplitude of the stride in trotters–comparison between fore-and hindlimbs, and influence of the track surface. Computer Methods in Biomechanics and Biomedical Engineering, 17(S1), 144-145. Pegasus (2017) Gaitsmart Equine Gait Analysis. Available from: http://www.gaitsmart.com/gaitsmart-pegasus/ [Accessed July 2017]. Peham, C., Scheidl, M. and Licka, T., (1996). A method of signal processing in motion analysis of the trotting horse. Journal of Biomechanics, 29(8), 1111-1114. Peham, C., Licka, T., Girtler, D. and Scheidl, M. (2001) Hindlimb lameness: clinical judgement versus computerised symmetry measurement. Veterinary Record, 148(24), 750-752. Persson, S. G. P., (1967) On blood volume and working capacity in horses. Acta Veterinaria Scandinavica, (S19), 9-189. Persson, S., (1969) Value of haemoglobin determination in the horse. Nordisk Veterinaer Medicin, 21(10), 513. Petrov, R., MacDonald, M. H., Tesch, A. M., Van Hoogmoed, L. M., (2003) Influence of topically applied cold treatment on core temperature and cell viability in equine superficial digital flexor tendons. American Journal of Veterinary Research, 64(7), 835-844. Pfau, T., Boultbee, H., Davis, H., Walker, A. and Rhodin, M., (2016a) Agreement between two inertial sensor gait analysis systems for lameness examinations in horses. Equine Veterinary Education. 28(4), 203-208.

http://avmajournals.avma.org/loi/ajvr

238

Pfau, T., Fiske-Jackson, A., and Rhodin, M., (2015) Quantitative assessment of gait parameters in horses: Useful for aiding clinical decision making? Equine Veterinary Education, 28(4), 209–215. Pfau, T., Jennings, C., Mitchell, H., Olsen, E., Walker, A., Egenvall, A., Tröster, S., Weller, R. and Rhodin, M., (2016b) Lungeing on hard and soft surfaces: Movement symmetry of trotting horses considered sound by their owners. Equine Veterinary Journal, 48(1), 83-89. Pfau, T., Robilliard, J. J., Weller, R., Jespers, K., Eliashar, E. and Wilson, A. M., (2007) Assessment of mild hindlimb lameness during over ground locomotion using linear discriminant analysis of inertial sensor data. Equine Veterinary Journal, 39(5), 407-413. Pfau, T., Stubbs, N.C., Kaiser, L.J., Brown, L.E. and Clayton, H.M., (2012) Effect of trotting speed and circle radius on movement symmetry in horses during lunging on a soft surface. American Journal of Veterinary Research, 73(12), 1890-1899. Pfau, T. and Weller, R., (2017) Comparison of a standalone consumer grade smartphone with a specialist inertial measurement unit for quantification of movement symmetry in the trotting horse. Equine Veterinary Journal, 49(1), 124-129. Pfau, T., Witte, T. H., and Wilson, A. M., (2005) A method for deriving displacement data during cyclical movement using an inertial sensor. The Journal of Experimental Biology, 208(13), 2503-2514. Pfau, T., Witte, T. H. and Wilson, A. M., (2006) Centre of mass movement and mechanical energy fluctuation during gallop locomotion in the Thoroughbred racehorse. Journal of Experimental Biology, 209(19), 3742-3757. Polefka, T. G., Bianchini, R. J., and Shapiro, S., (2012) Interaction of mineral salts with the skin: a literature survey. International Journal of Cosmetic Science, 34, 416-423. Pourcelot, P., Audigie, F., Degueurce, C., Denoix, J. M. and Geiger, D., (1997) Kinematic symmetry index: a method for quantifying the horse locomotion symmetry using kinematic data. Veterinary Research, 28(6), 525-538. Prankel, S., (2008) Hydrotherapy in Practice In Practice, 30(5), 272-277. Qualisys©; ProReflex, Qualisys© Ltd, Gothenburg, Sweden QTM (2015) Qualisys Track Manager User Manual. Available at: http://fy.chalmers.se/~f7xiz/TIF081C/QTM-usermanual.pdf [Accessed most recently: November 2015].

239

Ratzlaff, M. H., Hyde, M. L., Grant, B. D., Balch, O. and Wilson, P. D., (1990) Measurement of vertical forces and temporal components of the strides of horses using instrumented shoes. Journal of Equine Veterinary Science, 10(1), 23-35. Reesink, H. L., Divers, T. J., Bookbinder, L. C., van Eps, A. W., Soderholm, L. V., Mohammed, H. O., and Cheetham, J. (2012) Measurement of digital laminar and venous temperatures as a means of comparing three methods of topically applied cold treatment for digits of horses. American Journal of Veterinary Research, 73(6), 860. Rhodin, M., Álvarez, C. B., Byström, A., Johnston, C., Weeren, P. R., Roepstorff, L. and Weishaupt, M. A., (2009) The effect of different head and neck positions on the caudal back and hindlimb kinematics in the elite dressage horse at trot. Equine Veterinary Journal, 41(3), 274-279. Rhodin, M., Egenvall, A., Andersen, P.H. and Pfau, T., (2017) Head and pelvic movement asymmetries at trot in riding horses in training and perceived as free from lameness by the owner. PloS One, 12(4), p.e. 0176253. Rhodin, M., Johnston, C., Holm, K. R., Wennerstrand, J., and Drevemo, S., (2005) The influence of head and neck position on kinematics of the back in riding horses at the walk and trot. Equine Veterinary Journal, 37(1), 7-11. Rhodin, M., Pfau, T., Roepstorff, L. and Egenvall, A., (2013) Effect of lungeing on head and pelvic movement asymmetry in horses with induced lameness. The Veterinary Journal, 198, 39-45. Robert, C., Audigie, F., Valette, J. P., Pourcelot, P. and Denoix, J. M. (2001a) Effects of treadmill speed on the mechanics of the back in the trotting saddlehorse. Equine Veterinary Journal, 33(S33), 154-159. Robert, C., Valette, J. P. and Denoix, J. M., (2000) The effects of treadmill inclination and speed on the activity of two hindlimb muscles in the trotting horse. Equine Veterinary Journal, 32(4), 312-317. Robert, C., Valette, J. P., Denoix, J. M., (2001b) The effects of treadmill inclination and speed on the activity of three trunk muscles in the trotting horse. Equine Veterinary Journal, 33(5), 466-472. Robert, C., Valette, J. P., Pourcelot, P., Audigie, F., and Denoix, J. M., (2002) Effects of trotting speed on muscle activity and kinematics in saddlehorses. Equine Veterinary Journal, 34(S34), 295-301. Robert-Lachaine, X., Mecheri, H., Larue, C. and Plamondon, A., (2017) Validation of inertial measurement units with an optoelectronic system for whole-body motion analysis. Medical & Biological Engineering & Computing, 55(4), 609-619.

240

Robin, D., Chateau, H., Jacobs, F., Estoup, P., Holden, L., Falala, S., Pourcelot, P. and Crevier-Denoix, N., (2010) Suitability of a dynamometric horseshoe for the recording of the ground reaction forces on ridden horses. Computer Methods in Biomechanics and Biomedical Engineering, 13(S1), 129-131. Robilliard, J. J., Pfau, T. and Wilson, A. M. (2007) Gait characterisation and classification in horses Journal of Experimental Biology, 210(2), 187-197. Roepstorff, L. and Drevemo, S., (1993) Concept of a force-measuring horseshoe. Cells Tissues Organs, 146(2-3), 114-119. Roepstorff, L., Johnston, C., Drevemo, S., and Gustås, P., (2002). Influence of draw reins on ground reaction forces at the trot. Equine Veterinary Journal, 34(S34), 349-352. Roland, E. S., Hull, M. L. and Stover, S. M., (2005) Design and demonstration of a dynamometric horseshoe for measuring ground reaction loads of horses during racing conditions. Journal of Biomechanics, 38(10), 2102-2112. Saber-Sheikh, K., Bryant, E. C., Glazzard, C., Hamel, A. and Lee, R. Y., (2010) Feasibility of using inertial sensors to assess human movement. Manual Therapy, 15(1), 122-125. Scanlon, J., (1969) Animal Exercising, Conditioning and Therapy and Apparatus Therefor. United States Patent Office. Patent Number 3,485,213. Scott, R., Nankervis, K., Stringer, C., Westcott, K., and Marlin, D., (2010) The effect of water height on stride frequency, stride length and heart rate during water treadmill exercise. Equine Veterinary Journal, 42(S38), 662-664. Seite, S., (2013) Thermal waters as cosmeceuticals: La Roche-Posay thermal spring water example. Clinical, Cosmetic and Investigational Dermatology, 6, 23-28. Shapiro, S. S. and Wilk, M.B., (1965) An analysis of variance test for normality (complete samples). Biometrika, 52(3-4), 591-611. Singleton, W. H., (2001) Rein tension during horse-back riding activities. MS theses, Michigan State University. Slijper, E.J., (1946) Comparative Biologic-anatomical on the Vertebral column and Spinal musculature of Mammals investigations. Proceedings of the Koninklijke Nederlandse Akademie van Wetenschappen. 42, 1-128.

241

Starke, S. D., Willems, E., May, S. A. and Pfau, T., (2012a) Vertical head and trunk movement adaptations of sound horses trotting in a circle on a hard surface. The Veterinary Journal, 193(1), 73-80. Starke, S. D., Witte, T. H., May, S. A. and Pfau, T., (2012b) Accuracy and precision of hind limb foot contact timings of horses determined using a pelvis-mounted inertial measurement unit. Journal of Biomechanics, 45(8), 1522-1528. Strasser, H. (1913) Lehrbuch der Muskel- und Gelenkmechanik. (Cited in Jeffcott (1979) and cited in Sleijper (1946)) Sydney Free Saddlery (2017) Kavalkade HO Lungeing Aid. Available from: http://www.sydneyfree.co.uk/kavalkade-ho-lunging-aid.ir?cName=horse-saddlery-lunge-and-training-aids [Accessed July 2017] Tabor, G. and Williams, J., (2017) Equine Rehabilitation: A Review of Trunk and Hind Limb Muscle Activity and Exercise Selection. Journal of Equine Veterinary Science, 60, 97-103.

Thomas, D. P., Fregin, G. F., Gerber, N. H., and Ailes, N. B., (1980) Cardiorespiratory adjustments to tethered-swimming in the horse. Pflugers Archiv. European Journal of Physiology, 385(1), 65-70. Thomsen, M. H., Jensen, A. T., Sørensen, H., Lindegaard, C. and Andersen, P. H., (2010a) Symmetry indices based on accelerometric data in trotting horses. Journal of Biomechanics, 43(13), 2608-2612. Thomsen, M. H., Persson, A. B., Jensen, A. T., Sørensen, H. and Andersen, P. H., (2010b) Agreement between accelerometric symmetry scores and clinical lameness scores during experimentally induced transient distension of the metacarpophalangeal joint in horses. Equine Veterinary Journal, 42(S38), 510-515. Tokuriki, M., Ohtsuki, R., Kai, M., Hiraga, A., Oki, H., Miyahara, Y., and Aoki, O., (1999) EMG activity of the muscles of the neck and forelimbs during the different forms of locomotion. Equine Veterinary Journal, 31(S30), 231-234. Townsend, H. G. G., Leach, D. H. and Fretz, P. B., (1983) Kinematics of the equine thoracolumbar spine. Equine Veterinary Journal, 15(2), 117-122. Tranquille, C. A., Nankervis, K. J., Walker, V. A., Tacey, J. B. and Murray, R. C., (2017) Current knowledge of equine water treadmill exercise: what can we learn from human and canine studies? Journal of Equine Veterinary Science, 50, 76-83. Uhlir, C., Licka, T., Kübber, P., Peham, C., Scheidl, M. and Girtler, D., (1997) Compensatory movements of horses with a stance phase lameness. Equine Veterinary Journal, 29(S23), 102-105.

http://www.ncbi.nlm.nih.gov/pubmed?term=Thomas%20DP%5BAuthor%5D&cauthor=true&cauthor_uid=7191097

http://www.ncbi.nlm.nih.gov/pubmed?term=Fregin%20GF%5BAuthor%5D&cauthor=true&cauthor_uid=7191097

http://www.ncbi.nlm.nih.gov/pubmed?term=Gerber%20NH%5BAuthor%5D&cauthor=true&cauthor_uid=7191097

http://www.ncbi.nlm.nih.gov/pubmed?term=Ailes%20NB%5BAuthor%5D&cauthor=true&cauthor_uid=7191097

http://www.ncbi.nlm.nih.gov/pubmed?term=cardiorespiratory%20swimming%20horse

242

Ullman, D., (2009) The Curious Case of Charles Darwin and Homeopathy Advance Access Publication, 7(1), 33-39. Valentin, S., Davis, R., Wilson, A. and Pfau, T., (2010) The effect of loading on the equine spine - a preliminary study. Wiener Tierarztliche Monatsschrift, 97(3-4), 74-80. Van den Bogert, A. J., Van Weeren, P. R., and Schamhardt, H. C., (1990) Correction for skin displacement errors in movement analysis of the horse Journal of Biomechanics, 23(1), 97-101. Van de Winkel, D., de Bruijn, M., Touwen, N., Duchateau, L., Goethals, K., Oosterlinck, M., Pille, F., Van derperren, K., and Delesalle, C., (2016) Morphometric Changes in 15 Skeletal Muscles of Horses After 8 Weeks of Aquatraining Equine Veterinary Journal, 48(49), 30. Van Weeren, P. R. (2009) Kinematics of the equine back. In: Equine Back Pathology: Diagnosis and Treatment, Ed: F.M.D. Henson, Wiley-Blackwell, Oxford. pp 39-59. Van Weeren, P. R. and Barneveld, A., (1986) A technique to quantify skin displacement in the walking horse. Journal of Biomechanics, 19(10), 879-883. Van Weeren, P.R., Jansen, M.O., Van Den Bogert, A.J. and Barneveld, A., (1992a) A kinematic and strain gauge study of the reciprocal apparatus in the equine hind limb. Journal of Biomechanics, 25(11), 1291-1301. Van Weeren, P. R., McGowan, C. and Haussler, K. K., (2010) Science overview: Development of a structural and functional understanding of the equine back. Equine Veterinary Journal, 42(S38), 393-400. Van Weeren, P. R., Van Den Bogert, A. J., and Barneveld, A., (1990a) A quantitative analysis of skin displacement in the trotting horse. Equine Veterinary Journal, 22(S9), 101-109. Van Weeren P.R., Van Den Bogert, A.J., and Barneveld, A., (1990b) Quantification of skin displacement in the proximal parts of the limbs of the walking horse. Equine Veterinary Journal, 22(S9), 110-118. Van Weeren, P. R., Van Den Bogert, A. J., and Barneveld, A., (1992b) Correction models for skin displacement in equine kinematics gait analysis Journal of Equine Veterinary Science, 12(3), 178-192. Vincze, A., Szabó, C., Bakos, Z., Szabó, V., Veres, S., Ütő, D. and Hevesi, Á., (2016) Effect of dietary energy source on the plasma parameters of equine athletes trained in a deep water aqua treadmill. Italian Journal of Animal Science, 15(1), 137-143.

243

Voss, B., Mohr, E., and Krzywanek, H., (2002) Effects of Aqua-Treadmill Exercise on Selected Blood Parameters and on Heart-Rate Variability of Horses. Journal of Veterinary Medicine Series A, 49(3), 137-143. Waining, M., Young, I. S., Williams, S. B., (2011) Evaluation of the status of canine hydrotherapy in the UK. Veterinary Record, 168(15), 407-410. Walker, A. M., Wilson, A. M. and Pfau, T., (2010). Comparison of kinematic symmetry index calculations and the effects of straight and circular trotting. Equine Veterinary Journal, 42(S38), 482-487. Walker, V. A., Dyson, S. J. and Murray, R. C., (2013) Effect of a Pessoa training aid on temporal, linear and angular variables of the working trot The Veterinary Journal, 198(2), 404-411. Warner, S.M., Koch, T.O. and Pfau, T., (2010) Inertial sensors for assessment of back movement in horses during locomotion over ground. Equine Veterinary Journal, 42(S38), 417-424. Weishaupt, M. A., (2008) Adaptation strategies of horses with lameness. Veterinary Clinics of North America: Equine Practice, 24(1), 79-100. Weishaupt, M. A., Hogg, H. P., Wiestner, T., Demuth, D. C. and Auer, J. A., (1996) Development of a technique for measuring ground reaction force in the equine on a treadmill. In: Proceedings of the 3rd International Workshop on Animal Locomotion. p1. (Abstr.). Weishaupt, M.A., Wiestner, T., Hogg, H. P., Jordan, P. and Auer, J.A., (2004) Compensatory load redistribution of horses with induced weightbearing hindlimb lameness trotting on a treadmill. Equine Veterinary Journal, 36(8), 727-733. Weishaupt, M. A., Wiestner, T., Hogg, H. P., Jordan, P. and Auer, J. A., (2006) Compensatory load redistribution of horses with induced weight-bearing forelimb lameness trotting on a treadmill. The Veterinary Journal, 171(1), 135-146. Wennerstrand, J., Johnston, C., Roethlisberger-Holm, K., Erichsen, C., Eksell, P. and Drevemo, S., (2004) Kinematic evaluation of the back in the sport horse with back pain. Equine Veterinary Journal, 36(8), 707-711. Witte, T. H., Knill, K. and Wilson, A. M., (2004) Determination of peak vertical ground reaction force from duty factor in the horse (Equus caballus). Journal of Experimental Biology, 207(21), 3639-3648. Witte, T. H., Hirst, C. V. and Wilson, A. M., (2006) Effect of speed on stride parameters in racehorses at gallop in field conditions. Journal of Experimental Biology, 209(21), 4389-4397.

244

Witte, T. H. and Wilson, A.M., (2004) Accuracy of non-differential GPS for the determination of speed over ground. Journal of Biomechanics, 37(12), 1891-1898. Wyn-Jones, G. (1988) Equine Lameness, Blackwell Scientific Publications, Oxford. 6-7. Xing, H. Z., Zhang, Q. B., Braithwaite, C. H., Pan, B. and Zhao, J., (2017) High-speed photography and digital optical measurement techniques for geomaterials: fundamentals and applications. Rock Mechanics and Rock Engineering, 50(6), 1611-1659. Xsens©, Xsens© Ltd, Enschede, The Netherlands Yarnell, K., Fleming, J., Stratton, T. D., Brassington, R., (2014) Monitoring changes in skin temperature associated with exercise in horses on a water treadmill by use of infrared thermography. Journal of Thermal Biology, 45, 110-116. Zsoldos, R. R., Kotschwar, A. B., Kotschwar, A., Groesel, M., Licka, T. and Peham, C., (2010) Electromyography activity of the equine splenius muscle and neck kinematics during walk and trot on the treadmill. Equine Veterinary Journal, 42(S38), 455-461.

Kinematics of the equine axial skeleton -...

Documents

Transcript of Kinematics of the equine axial skeleton -...